StreetEYE Blog

Why Are There Recessions? Going On Tilt About Business Cycles

rena-listing_2023721i

To everything, turn, turn, turn.
There is a season, turn, turn, turn.
And a time to every purpose under heaven.
– The Byrds, by way of Ecclesiastes

Noah Smith has a good post on recessions, in which he highlights the importance of sticky prices, and the academic ‘debate’ over sticky prices. I’m not sure there has really ever been a legitimate debate over the existence of sticky prices and wages.

But his focus on a proximate cause omits an elephant in the room. Cycles are pervasive in nature and in economic activity. Why is that? What makes a process cyclical, and what makes the economy more or less prone to large cyclical swings?

Cycles in science

Take chemistry, as a random field of study. If you take one solution and add another, one of three things can happen.

  • Nothing. A stable, still inert solution.
  • A big explosion! A self-sustaining reaction and you move to a new stable state. (It doesn’t need to be an explosion, could be any self-sustaining, entropy favorable reaction, gas or crystals forming, etc., some sort of reaction and possibly phase change, culminating in a new stable state.)  

  • An oscillation! – A bit rare in chemistry but here is a weird example:  

The same dynamics apply to any system: Disturb a stable system, or organism, population, or environment, and one of those three things will happen.

Take a pendulum. Move it a (small) distance off center. There is a force pushing it back toward the center, which is proportional to how far off center it is. The force accelerates the ball back toward the center. Accelerates means it speeds up over time, which means speed responds with a time lag to how far off center the pendulum is. When it gets back to the neutral position, it still has speed, and at neutral no force keeps it in the center, so it overshoots. Then the process starts back in the opposite direction.

The resulting position, acceleration, and velocity of the pendulum are each described by sine waves.

And the exact same math describes any system with the same dynamics. A force proportional to distance from neutral back in the neutral direction, and accelerating velocity proportional to that force. It works for springs, it works for some electrical circuits, it works for ocean waves. Same math always gives the exact same sine waves.

Cycles in business activity

Now, let’s look at an industry where there is constant growth of say 10%, and producers seek to maintain 10% of annual sales as year-end inventory, so they can survive 1/10th of the year in the event of a supply interruption. In steady state growth, orders, sales and inventories rise in lockstep at 10%.

Inventory cycle - base

Now demand growth unexpectedly increases to 20%. Distributors initially supply the extra demand out of their inventory, and increase orders for the following year. The problem is, they are now 10% short of inventory. So they need to order enough to 1) supply customers at the higher growth rate, 2) restore the inventory they just depleted, and 3) grow inventories to support the new level of sales. So orders will go up a lot more than the ‘new normal’ 20%, more like 40%.

Inventory cycle - boom
Now suppose production, orders, and inventory are growing steadily at 20% and suddenly go flat. At the end of the year you have ~20% too much inventory. For the following year your orders will not be flat, they will drop 20%.

Inventory cycle - bust

The point is, the inventory cycle is like the pendulum: when there is a shock, activity is out of balance, there is a force returning to balance with a lag, and over-correction in the opposite direction.

Procyclical and countercyclical dynamics

The economy and markets are loaded with procyclical dynamics which make growth or recession self-reinforcing, and countercyclical dynamics which dampen business cycle swings.

Procyclical:

Inventories, as we just discussed.

Capital equipment. When you invest in plant and equipment in anticipation of growth, and the growth doesn’t materialize, you have too much capital equipment. Your orders for capital equipment don’t just decline, they go to zero. Capital equipment is like an inventory of future production. They call these industries and stocks cyclical for a reason.

Consumer debt. Suppose you spend all your income, and you have consumer debt, and you maintain the principal outstanding at 50% of income. When your income goes up 10%, you spend 10% more, and you also increase your debt 10%, allowing you to spend 15% more every year. When your income suddenly goes down, you have to reduce your spending growth in line with income and reduce your debt to be line with income, leading to an even larger drop in spending.

Government spending. Can have the same dynamics as consumer spending when the government targets a debt/GDP ratio. In the case of a deficit target, like the euro zone’s stability and growth pact, when recession hits, taxes drop and spending on social services goes up. To maintain a fixed deficit/GDP target in the face of declining GDP and expanding cyclical deficit, you have to enact procyclical austerity, cutting spending in the teeth of a recession. The same logic applies to a balanced budget amendment for the USA.

Bubble dynamics in asset markets. Price rises and declines can be self-fulfilling. If you need a house and prices are going down, you’re in no rush to buy. As soon as the market starts climbing, everyone who was on the sidelines is in a rush to buy before prices rise further, and sellers are no longer in a rush to sell. That’s why, as soon as there is a sense that ‘the bottom is in,’ there is an immediate sharp turnaround. And of course, the trend feeds into consumption and the overall economy. Rising stock prices and house prices mean more people spend, buy more houses, which creates construction and more income and profits, which leads to more demand and higher prices for houses and stocks.

Conversely financial crisis dynamics mean that if one institution goes belly up, you don’t know what other institution might have just lost money in the failure and itself be endangered. Investors sell everything, and you potentially have a run on the market. The more levered and opaque the financial system is, the greater the fragility and risk of domino failures. In a Fisher debt-deflation cycle, which the Fed has been extremely keen to avoid post-2008, overly indebted consumers stop spending, which leads to deflation, which increases the real value of their debt, which leads to bankruptcies and even more of a demand decline, perpetuating the cycle.

Any other process that requires a prediction, and possibly belated over-adjustment if the prediction turns out wrong.

Anything that increases multiplier effects. When people spend more of an increase in income, multiplier effects of an income shock are greater, when they save more the fiscal multiplier is smaller. When long-term interest rates are near zero, asset prices and wealth are more sensitive to cash flows far out in the future. Kind of like growth stocks are high beta because they are valuing big earnings far in the future, and a small swing in required return or anticipated growth rate has a big impact on those future numbers.

Countercyclical:

Anything that reduces inventory and capex investment swings: Faster information diffusion about changes in demand, production, and inventories; information technology that lets firms quickly adjust forecasts and orders; just-in-time inventory systems; increasing the mix of industries in the economy that are less capital- and inventory-intensive (notably services).

Automatic stabilizers. Automatic increases in government spending on unemployment insurance, social security allow government spending and deficits to automatically rise when the economy slows.

The economy’s inherent stability: high demand leads to higher prices, which reduce demand; falling demand leads to lower prices, which restore demand. The faster prices and wages adjust, the faster people can move to new locales and new industries for new jobs, the lower the shock to GDP. Hence sticky prices and wages contribute to instability and recessions. This is why the ECB is always going on about structural reform.

Discretionary countercyclical fiscal policy, like the Bush rebates (everyone’s a Keynesian in an election year).

Countercyclical monetary policy, raising interest rates as the economy approaches full employment, whether discretionary or via something like the Taylor rule.

Concluding comments

GDP doesn’t look like a sine wave. It grows and fluctuates around trend population growth plus productivity growth. It seems to fluctuate randomly around the trend, and then occasionally something breaks bad, kicking off a sharp self-reinforcing contractionary cycle — a recession — followed by a gradual return to trend.

There are a lot of causes for recessions. Supply shock (e.g. oil crisis). Demand shock, e.g. financial crisis where a lot of perceived wealth evaporates. For instance, 2008, the tulip bubble, the South Sea bubble, the Mississippi bubble, the 1825 crisis.

A good analogy is a boat – tip it a little to starboard and it rights itself. If you have a heavy keel it will be more stable. If you pile heavy cargo and fuel tanks high on deck, don’t secure things so they can’t shift from side to side, it capsizes when it reaches a tipping point that overcomes its limited inherent stability.

Sticky prices and wages are one reason why economic adjustments lag and take the form of business cycles and recessions. They are readily observed, and readily explained based on loss aversion. But cycles are everywhere any dynamic operates with a lag and leads to overshoot. Sticky prices are one heavy stack of containers towering above deck, but financial markets, real estate, capex boom-bust cycles, and policy can all play a role.

There is legitimate debate over whether humans are smart enough to apply discretionary fiscal and monetary policy, or if a rules-based process like a Taylor rule for monetary policy, and automatic stabilization built into fiscal policy, is more effective than a sluggish, political, and not always all-wise discretionary policy-making process.

There are strains of thought that say the economy is self-stabilizing, and all recessions are caused by government policy interference with otherwise perfect decision-makers. That’s like saying the Titanic is unsinkable. However big the boat, it obeys the same laws of physics. An experience like 2008 shows we can’t model all the things that cause instability, and we can’t assume everyone is making perfect decisions. Given a big enough hole in the hull, nothing is unsinkable, and standing around assuming it’s going to fix itself is not always the best course.

Opinions can differ over the efficacy of stabilization policy. But if your model that says recessions and business cycles shouldn’t happen, it’s not very useful for understanding them.

Successful investors like Keynes and George Soros were alert to instability. Equilibrium is like efficient markets – a useful assumption for a simple model, that works great in theory but not always in practice.

Talk of the Town – Benzinga Fintech Awards #BZawards

A lot of disruption on display at Wednesday’s Benzinga Fintech awards. Increasingly, institutional-quality platforms available across Web and mobile at ultra-low price points, and new information diffusion networks and investing foodchains via social and crowd-sourcing.

  • Vestorly – a one-stop shop for an advisor to manage social media presence, content marketing, lead generation. Sort of a combination of Hootsuite and a website live news widget. From the same app, you can update all your social media accounts (Hootsuite type functionality), and update your website with a widget containing latest live news you want to share. It’s an interesting idea…although just using Hootsuite to update all your social media accounts, and then putting e.g. a Twitter widget on your website seems like a pretty good option.
  • NewsHedge – Web-based audio squawk that alerts you when strange things are afoot in markets.
  • Estimize – pushing to become the gold standard in consensus earnings estimates, economic forecasts, etc. (just announced their B funding round)
  • Market Prophit – Sentiment analysis for social media mentions of stocks. Now, I got to be honest, I have no idea if it actually predicts market performance. But somebody has to try it.
  • A bunch of ultra-low-cost investing platforms – Motif, Betterment
  • Platforms to let investors use sophisticated institutional strategies – Quantopian, Crowdfunding like Circleup, although honestly Angellist seems like the axe.
  • Etna – So, I trade with Interactive Brokers, and they provide TWS, the trader workstation app written in Java that gives real-time data, charts, trading, portfolio analytics, etc. Etna offers a platform with similar functionality that online brokers can provide their clients. Except it’s written in HTML5 and Javascript, it’s just an interactive website. And it looks pretty amazing. If it works as well as it looks, that’s the way of the future. No more downloading Java updates and TWS updates.
  • If you use technical charts and real-time charts, ChartIQ is worth a look. I did the free trial a while back and it’s pretty powerful. But I gotta admit, I’m pretty old school, so I went back to StockCharts.com. I have it set up as a Chrome search engine, so I type “chart NFLX” and I get a chart with all my custom options. I don’t really need real-time updating charts or anything too complicated, I just want a simple chart exactly the way I want it.

ReformedBroker – The mayor of the financial twittersphere brought down the house.
SallyPancakes
Linette Lopez
Greg Neufeld Twitter
Will Ortel from CFA Institute
Elliot Spitzer – who wasn’t seen asking all the startups how their technology could be applied to meeting women, but is actually an investor and board member of some fintech startups.

I definitely missed some folks and some great companies.

We get a little jaded but there really is an amazing amount of disruption going on right now. And a lot of inspiring and awesome people making it happen. Big thanks to Jason Raznick and Kyle Bazzy of Benzinga for putting it all together.

Gold as Part of a Long-Run Asset Allocation (update)

You have to choose between trusting to the natural stability of gold and the natural stability of the honesty and intelligence of the members of the government. And, with due respect to these gentlemen, I advise you, as long as the capitalist system lasts, to vote for gold. – George Bernard Shaw

Here’s a quick update of a post I did a couple of years back on Gold as part of a long run asset allocation. Gold hasn’t fared too well since then.

Let’s look at four asset classes from 1928-2014: US stocks (ie S&P), medium-term Treasurys (ie 10-year), T-bills, and gold. (Would love to do international developed, emerging, TIPS, real estate, but data doesn’t go back that far.)

Let’s adjust returns for inflation. Here’s are the historical mean annual real returns and standard deviations of annual returns.

Real Return Real Risk
Stocks 8.3% 19.8%
Bonds 2.3% 8.8%
Bills 0.5% 3.9%
Gold 3.2% 18.8%

Let’s compute the efficient frontier. The left-most point is the minimum-volatility portfolio. The right-most point is the max-return portfolio, which is 100% stocks. We compute the minimum-volatility portfolio for return levels between those two, and plot the resulting efficient frontier.

Efficient Frontier, 1928-2014

What is the composition of the portfolio at each point on the efficient frontier? We plot a transition map showing that as you start from the minimum-volatility portfolio with about 1% real return and 2% volatility, composed of mostly T-bills, with some stocks and gold, and move toward the maximum-return portfolio, you add more and more stocks, but always include some gold.

Transition map, 1926-2014
Transition map

Let’s try a few different eras.

1946-2014, Post-war, since Bretton Woods:

Efficient frontier

Transition map
 

1972-2010, Post-war, post-gold standard (had to adjust the scale a little to get that gold data point on there):

Efficient frontier

Transition map
 

1982-2014, era of disinflation:

Efficient frontier

Transition map

What should one conclude? In most regimes gold was worth owning in the portfolio that gives the most return at a given risk level. The exception was the era of globalization and disinflation, where we had high returns from stocks coupled with disinflation. If you expect that to be the case, as it has been the last 30 years, gold doesn’t improve the longer time-frame, more risky portfolios, like a 70-30 portfolio. But over the varied regimes of the last 87 years, it was a hedge worth having.

I say this as one who believes the gold bugs are useless, except for a chuckle. But central banks really want moderate inflation to solve the consumer debt/balance sheet problem. Deflation is anathema to them when everyone is up to their eyeballs in debt.

The question of our time is whether QE/easing -> inflated asset values -> more debt -> consumer goods/services inflation -> solves debt and overinflated asset problem.

Or QE/easing -> more debt -> deflation/no inflation -> even more precarious balance sheets -> financial crises and economic chaos.

Either way, a little gold is a good hedge in a number of scenarios.

(See the whole Bernanke/Summers/Piketty secular stagnation/robots debate, which I discussed a bit here.)

R code and data:

?View Code RSPLUS
# install.packages('quantmod')
# require(quantmod)
# install.packages('lpSolve')
require(lpSolve)
# install.packages('quadprog')
require(quadprog)
# install.packages('ggplot2')
require(ggplot2)
require(reshape)
 
# define functions
 
#################################################################
# use linear programming to find maximum return portfolio (100% highest return asset)
#################################################################
 
runlp <- function ( returns )
{
 
	# find maximum return portfolio (rightmost point of efficient frontier)
	# will be 100% of highest return asset
	# maximize
	#   w1 * stocks return +w2 *bills +w3*bonds + w4 * gold
	#   subject to 0 <= w <= 1  for each w
	# will pick highest return asset with w=1
	# skipping >0 constraint, no negative return assets, so not binding
 
	opt.objective <- apply(returns, 2, mean)
 
	# should use length(objective) to populate matrix
	nAssets <- length(returns)
	ones = rep (1, nAssets)
	zeros = rep (0, nAssets)
 
	# constrain sum of weights to 1
	constraintlist = ones
	operatorlist = c("=")
	rhslist = c(1)
 
	# constrain each weight >= 0
	for(i in 1:nAssets) {
		newconstraint = zeros
		newconstraint[i]=1
		constraintlist = c(constraintlist, newconstraint)
		operatorlist = c(operatorlist, ">=")
		rhslist = c(rhslist, 0)
	}
 
#	Example
#	opt.constraints <- matrix (c(1, 1, 1, 1,  # constrain sum of weights to 1
#							 1, 0, 0, 0,  # constrain w1 <= 1
#							 0, 1, 0, 0,  # constrain w2 <= 1
#							 0, 0, 1, 0,  # constrain w3 <= 1
#							 0, 0, 0, 1)  # constrain w4 <= 1
#						   , nrow=5, byrow=TRUE)
 
	opt.constraints <- matrix (constraintlist, nrow=nAssets+1, byrow=TRUE)
	opt.operator <- operatorlist
	opt.rhs <- rhslist
	opt.dir="max"
 
	tmpsolution = lp (direction = opt.dir,
	opt.objective,
	opt.constraints,
	opt.operator,
	opt.rhs)
 
	sol= c()
	# portfolio weights for max return portfolio
	sol$wts=tmpsolution$solution
	# return for max return portfolio
	sol$ret=tmpsolution$objval
	# compute return covariance matrix to determine volatility of this portfolio
	sol$covmatrix = cov(returns, use = 'complete.obs', method = 'pearson')
	# multiply weights x covariances x weights, gives variance
	sol$var = sol$wts %*% sol$covmatrix %*% sol$wts
	# square root gives standard deviation (volatility)
	sol$vol = sqrt(sol$var)
 
	return (sol)
}
 
runqp <- function ( returns, hurdle=0 )
{
#################################################################
# find minimum volatility portfolio
#################################################################
 
# minimize variance:  w %*% covmatrix %*% t(w)
# subject to sum of ws = 1
# subject to each w >= 0
# subject to each return >= hurdle
 
# solution.minvol <- solve.QP(covmatrix, zeros, t(opt.constraints), opt.rhs, meq = opt.meq)
# first 2 parameters covmatrix, zeros define function to be minimized
# if zeros is all 0s, the function minimized ends up equal to port variance / 2
# opt.constraints is the left hand side of the constraints, ie the cs in
# c1 w1 + c2 w2 ... + cn wn = K
# opt.rhs is the Ks in the above equation
# meq means the first meq rows are 'equals' constraints, remainder are >= constraints
# if you want to do a <= constraint, multiply by -1 to make it a >= constraint
# does not appear to accept 0 RHS, so we make it a tiny number> 0
 
	# compute expected returns
	meanreturns <- apply(returns, 2, mean)
 
	# compute covariance matrix
	covmatrix = cov(returns, use = 'complete.obs', method = 'pearson')
 
	nAssets <- length(returns)
	nObs <- length(returns$stocks)
	ones = rep (1, nAssets)
	zeros = rep (0, nAssets)
 
	# constrain sum of weights to 1
	constraintlist = ones
	rhslist = c(1)
 
	# constrain each weight >= 0
	for(i in 1:nAssets) {
		newconstraint = zeros
		newconstraint[i]=1
		constraintlist = c(constraintlist, newconstraint)
		rhslist = c(rhslist, 0)
	}
 
	# constrain return >= hurdle
	constraintlist = c(constraintlist, meanreturns)
	rhslist = c(rhslist, hurdle)
 
	# example
	# opt.constraints <- matrix (c(1, 1, 1, 1,   # sum of weights =1
	#							 1, 0, 0, 0,   # w1 >= 0
	#							 0, 1, 0, 0,   # w2 >= 0
	#							 0, 0, 1, 0,   # w3 >= 0
	#							 0, 0, 0, 1)   # w4 >= 0
 
	#						   , nrow=5, byrow=TRUE)
	# opt.rhs <- matrix(c(1, 0.000001, 0.000001, 0.000001, 0.000001))
	# opt.constraints = rbind(opt.constraints, meanreturns)
	# opt.rhs=rbind(opt.rhs, hurdle)
 
	opt.constraints <- matrix (constraintlist, nrow=nAssets+2, byrow=TRUE)
	opt.rhs <- opt.rhs <- matrix(rhslist)
	opt.meq <- 1  # first constraint is '=', rest are '>='
 
	zeros <- array(0, dim = c(nAssets,1))
	tmpsolution <- solve.QP(covmatrix, zeros, t(opt.constraints), opt.rhs, meq = opt.meq)
 
	sol= c()
	sol$wts = tmpsolution$solution
	sol$var = tmpsolution$value *2
	sol$ret = meanreturns %*% sol$wts
	sol$vol = sqrt(sol$var)
 
	return(sol)
}
 
loopqp <- function (minvol, maxret, numtrials)
{
 
	#################################################################
	# loop and run a minimum volatility optimization for each return level from 2-49
	#################################################################
 
	# put minreturn portfolio in return series for min return, index =1
	out.ret=c(minvol$ret)
	out.vol=c(minvol$vol)
	out.stocks=c(minvol$wts[1])
	out.bills=c(minvol$wts[2])
	out.bonds=c(minvol$wts[3])
	out.gold=c(minvol$wts[4])
 
	lowreturn <- minvol$ret
	highreturn <- maxret$ret
	minreturns <- seq(lowreturn, highreturn, length.out=numtrials)
 
	for(i in 2:(length(minreturns) - 1)) {
		tmpsol <- runqp(freal,minreturns[i])
		tmp.wts = tmpsol$wts
		tmp.var = tmpsol$var
 
		out.ret[i] = realreturns %*% tmp.wts
		out.vol[i] = sqrt(tmp.var)
		out.stocks[i]=tmp.wts[1]
		out.bills[i]=tmp.wts[2]
		out.bonds[i]=tmp.wts[3]
		out.gold[i]=tmp.wts[4]
	}
 
# put maxreturn portfolio in return series for max return
	out.ret[numtrials]=c(maxret$ret)
	out.vol[numtrials]=c(maxret$vol)
	out.stocks[numtrials]=c(maxret$wts[1])
	out.bills[numtrials]=c(maxret$wts[2])
	out.bonds[numtrials]=c(maxret$wts[3])
	out.gold[numtrials]=c(maxret$wts[4])
 
	efrontier=data.frame(out.ret*100)
	efrontier$vol=out.vol*100
	efrontier$stocks=out.stocks*100
	efrontier$bills=out.bills*100
	efrontier$bonds=out.bonds*100
	efrontier$gold=out.gold*100
	names(efrontier) = c("Return", "Risk", "%Stocks", "%Bills", "%Bonds", "%Gold")
 
	return(efrontier)
}
 
############################################################
# charts
############################################################
 
plot_efrontier <- function (efrontier, returns, sds, apoints, title) {
 
     ggplot(data=efrontier, aes(x=Risk, y=Return)) +
          theme_bw() +
	  geom_line(size=1.4) +
	  geom_point(data=apoints, aes(x=Risk, y=Return)) +		
	  scale_x_continuous(limits=c(1,24)) +
	  ggtitle(title) +
	  annotate("text", apoints[1,1], apoints[1,2],label=" stocks", hjust=0) +
	  annotate("text", apoints[2,1], apoints[2,2],label=" bills", hjust=0) +
	  annotate("text", apoints[3,1], apoints[3,2],label=" bonds", hjust=0) +
	  annotate("text", apoints[4,1], apoints[4,2],label=" gold", hjust=0) +
	  annotate("text", 19,0.3,label="streeteye.com", hjust=0, alpha=0.5)
 
}
 
plot_transitionmap <- function (efrontier, returns, sds) {
 
	# define colors
	dvblue = "#000099"
	dvred = "#e41a1c"
	dvgreen = "#4daf4a"
	dvpurple = "#984ea3"
	dvorange = "#ff7f00"
	dvyellow = "#ffff33"
	dvgray="#666666"
 
	efrontier.m = melt(efrontier, id ='Risk')
 
	ggplot(data=efrontier.m, aes(x=Risk, y=value, colour=variable, fill=variable)) +
		theme_bw() +
		theme(legend.position="top", legend.direction="horizontal") +
		ylab('% Portfolio') +
		geom_area() +
		scale_colour_manual("", breaks=c("%Stocks", "%Bills", "%Bonds","%Gold"), values = c(dvblue,dvgreen,dvred,dvyellow), labels=c('%Stocks', '%Bills','%Bonds','%Gold')) +
		scale_fill_manual("", breaks=c("%Stocks", "%Bills", "%Bonds","%Gold"), values = c(dvblue,dvgreen,dvred,dvyellow), labels=c('%Stocks', '%Bills','%Bonds','%Gold'))
#		annotate("text", 16,-2.5,label="streeteye.com", hjust=0, alpha=0.5)
 
}
 
#################################################################
# Create some data
#################################################################
# sources:
# http://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
# http://www.onlygold.com/Info/Historical-Gold-Prices.asp
# http://www.spdrgoldshares.com/usa/historical-data/
#################################################################
 
# not used in abbreviated example, but useful for reporting
startYear = 1928
endYear = 2014
YEARS =startYear:endYear
 
# nominal returns
# nominal returns
SP500 = c(0.4381,-0.083,-0.2512,-0.4384,-0.0864,0.4998,-0.0119,0.4674,0.3194,-0.3534,0.2928,-0.011,
-0.1067,-0.1277,0.1917,0.2506,0.1903,0.3582,-0.0843,0.052,0.057,0.183,0.3081,0.2368,0.1815,
-0.0121,0.5256,0.326,0.0744,-0.1046,0.4372,0.1206,0.0034,0.2664,-0.0881,0.2261,0.1642,0.124,
-0.0997,0.238,0.1081,-0.0824,0.0356,0.1422,0.1876,-0.1431,-0.259,0.37,0.2383,-0.0698,0.0651,
0.1852,0.3174,-0.047,0.2042,0.2234,0.0615,0.3124,0.1849,0.0581,0.1654,0.3148,-0.0306,0.3023,
0.0749,0.0997,0.0133,0.372,0.2268,0.331,0.2834,0.2089,-0.0903,-0.1185,-0.2197,0.2836,0.1074,
0.0483,0.1561,0.0548,-0.3655,0.2594,0.1482,0.021,0.1589,0.3215,0.1348)
 
BILLS = c(0.0308,0.0316,0.0455,0.0231,0.0107,0.0096,0.0032,0.0018,0.0017,0.003,0.0008,0.0004,
0.0003,0.0008,0.0034,0.0038,0.0038,0.0038,0.0038,0.0057,0.0102,0.011,0.0117,0.0148,
0.0167,0.0189,0.0096,0.0166,0.0256,0.0323,0.0178,0.0326,0.0305,0.0227,0.0278,0.0311,
0.0351,0.039,0.0484,0.0433,0.0526,0.0656,0.0669,0.0454,0.0395,0.0673,0.0778,0.0599,
0.0497,0.0513,0.0693,0.0994,0.1122,0.143,0.1101,0.0845,0.0961,0.0749,0.0604,0.0572,
0.0645,0.0811,0.0755,0.0561,0.0341,0.0298,0.0399,0.0552,0.0502,0.0505,0.0473,0.0451,
0.0576,0.0367,0.0166,0.0103,0.0123,0.0301,0.0468,0.0464,0.0159,0.0014,0.0013,0.0003,
0.0005,0.0007,0.0005)
 
BONDS=c(0.0084,0.042,0.0454,-0.0256,0.0879,0.0186,0.0796,0.0447,0.0502,0.0138,0.0421,0.0441,
0.054,-0.0202,0.0229,0.0249,0.0258,0.038,0.0313,0.0092,0.0195,0.0466,0.0043,-0.003,
0.0227,0.0414,0.0329,-0.0134,-0.0226,0.068,-0.021,-0.0265,0.1164,0.0206,0.0569,0.0168,
0.0373,0.0072,0.0291,-0.0158,0.0327,-0.0501,0.1675,0.0979,0.0282,0.0366,0.0199,0.0361,
0.1598,0.0129,-0.0078,0.0067,-0.0299,0.082,0.3281,0.032,0.1373,0.2571,0.2428,-0.0496,
0.0822,0.1769,0.0624,0.15,0.0936,0.1421,-0.0804,0.2348,0.0143,0.0994,0.1492,-0.0825,
0.1666,0.0557,0.1512,0.0038,0.0449,0.0287,0.0196,0.1021,0.201,-0.1112,0.0846,0.1604,
0.0297,-0.091,0.1075)
 
CPI=c(-0.0115607,0.005848,-0.0639535,-0.0931677,-0.1027397,0.0076336,0.0151515,0.0298507,
0.0144928,0.0285714,-0.0277778,0,0.0071429,0.0992908,0.0903226,0.0295858,0.0229885,
0.0224719,0.1813187,0.0883721,0.0299145,-0.0207469,0.059322,0.06,0.0075472,0.0074906,
-0.0074349,0.0037453,0.0298507,0.0289855,0.0176056,0.017301,0.0136054,0.0067114,0.0133333,
0.0164474,0.0097087,0.0192308,0.0345912,0.0303951,0.0471976,0.0619718,0.0557029,0.0326633,
0.0340633,0.0870588,0.1233766,0.0693642,0.0486486,0.0670103,0.0901771,0.1329394,0.125163,
0.0892236,0.0382979,0.0379098,0.0394867,0.0379867,0.010979,0.0443439,0.0441941,0.046473,
0.0610626,0.0306428,0.0290065,0.0274841,0.026749,0.0253841,0.0332248,0.017024,0.016119,
0.0268456,0.0338681,0.0155172,0.0237691,0.0187949,0.0325556,0.0341566,0.0254065,0.0408127,
0.0009141,0.0272133,0.0149572,0.03,0.017,0.015,0.008)
 
GOLD = c(0,0,0,0,0,0.563618771,0.082920792,
0,0,0,0,0,-0.014285714,0.028985507,0,
0.028169014,-0.006849315,0.027586207,0.026845638,0.124183007,-0.023255814,-0.035714286,
-0.00617284,-0.00621118,-0.0325,-0.082687339,-0.007042254,-0.002836879,0.001422475,
0.001420455,0,0,0.035460993,-0.02739726,-0.004225352,-0.002828854,
0.002836879,0.004243281,-0.002816901,0.002824859,0.225352113,-0.057471264,-0.051219512,
0.146529563,0.431390135,0.667919799,0.725864012,-0.242041683,-0.03962955,0.204305898,
0.291744258,1.205670351,0.296078431,-0.327618087,0.1175,-0.149888143,-0.189473684,
0.061688312,0.195412844,0.244563827,-0.156937307,-0.022308911,-0.036907731,-0.085577421,
-0.057057907,0.176426426,-0.021697511,0.009784736,-0.046511628,-0.222086721,0.005748128,
0.005368895,-0.060637382,0.014120668,0.23960217,0.217359592,0.04397843,0.17768595,
0.231968811,0.319224684,0.043178411,0.250359299,0.292413793,0.089292067,0.082625735,
-0.273303167,0.00124533
)
 
# truncate here, e.g.
# 1928 - 2014 - 87 years
# 1946 - 2014 - 69 years
 
#SP500=SP500[19:87]
#BILLS=BILLS[19:87]
#BONDS=BONDS[19:87]
#GOLD=GOLD[19:87]
#CPI=CPI[19:87]
 
# 1972 - 2014 - 43 years
# SP500=SP500[45:87]
# BILLS=BILLS[45:87]
# BONDS=BONDS[45:87]
# GOLD=GOLD[45:87]
# CPI=CPI[45:87]
 
# 1982 - 2014 - 33 years
SP500=SP500[55:87]
BILLS=BILLS[55:87]
BONDS=BONDS[55:87]
GOLD=GOLD[55:87]
CPI=CPI[55:87]
 
 
# put into a data frame
fnominal=data.frame(stocks=SP500, bills=BILLS, bonds=BONDS, gold=GOLD, CPI=CPI)
freal=data.frame(stocks=(1+SP500)/(1+CPI)-1, bills=(1+BILLS)/(1+CPI)-1, bonds=(1+BONDS)/(1+CPI)-1, gold=(1+GOLD)/(1+CPI)-1)
#freal=data.frame(stocks=SP500-CPI, bills=BILLS-CPI, bonds=BONDS-CPI, gold=GOLD-CPI)
 
# compute real return means
realreturns = apply(freal, 2, mean)
realreturnspct = realreturns*100
# print them
realreturnspct
 
# compute real return volatility (standard deviation of real returns)
realsds = apply(freal, 2, sd)
realsdspct = realsds*100
# print them
realsdspct
 
maxret <- runlp(freal)
minvol <- runqp(freal,0)
 
# generate a sequence of 50 evenly spaced returns between min var return and max return
efrontier = loopqp(minvol, maxret, 50)
 
apoints <- data.frame(realsdspct)
apoints$returns <- realreturnspct
names(apoints) = c("Risk", "Return")
 
plot_efrontier(efrontier, realreturnspct, realsdspct, apoints, "Efficient Frontier, 1946-2014")
keep=c("Risk", "%Stocks","%Bills","%Bonds","%Gold")
plot_transitionmap(efrontier[keep], realreturnspct, realsdspct)

Good risks and bad risks

4184728-16x9-940x529

Matthias Steiner, Beijing 2008

Pain is weakness leaving the body, and/or your central nervous system telling you you’re about to die. – seen on T-shirt

No matter what kind of math you use, you wind up measuring volatility with your gut. – Ed Seykota

Never tell me the odds. – Han Solo

The difference between a good risk and bad risk is sort of like the difference between good pain and bad pain when you’re working out.

Good pain: You’re squatting your personal record and every fiber of your being is saying drop it, and your head is exploding and you’re making weird grunting noises and you just might vomit or soil yourself…but you keep going for that one last rep with correct form and you feel major burnout and yet you feel great, because you know that is the burn that means progress. (I hate squats.)

Bad pain: You feel a little off today and you’re just going through the motions and you jerk it, instead of that perfect form you were taught. And then something in your back tightens up, and you get that sudden feeling, uh-oh, I might not be able to tie my shoes tomorrow.

No pain no gain. What doesn’t kill you makes you stronger, as Nietzsche said, shortly before he died.

But how do you learn the difference between the pain that makes you stronger and the pain that might kill you?

In Band of Brothers, on the DVD extras, Carwood Lipton talked about attacking the guns at Brécourt Manor with 12 soldiers under Dick Winters, vs. about 60 Germans.

That seems like bad odds to start with. (Allegedly, HQ ordered Dick Winters to take out the guns, believing he had hooked up with most of his still-scattered Easy Company).

Winters ordered Lipton to lay down covering fire, and he climbed a tree and started shooting down at the Germans in the trenches. Cleverly isolating the guns and storming them one by one, the small group took out the gun battery with minimal casualties.

But the older and wiser real-life Lipton interviewed 50 years later for the DVD said that later in the war he would never have climbed that tree, he was far too exposed and just didn’t know any better on his first day in combat.

Canadian Tommy Prince was in an Italian farmhouse as an observer to direct shelling, when a shellburst cut his phone line. He put on the Italian farmer’s clothes, went out like a farmer and inspected the chicken coop, and shook a fist at the Germans and the Allies. Then he leaned down as if to tie his shoelace, spliced the wire, and went back to directing fire on the Germans. Incredibly bold. But if you think about it, a tactic that probably reduced his risk profile vs. hunkering down incommunicado or making a run for it.

Bad risk: You keep doing that, you’re going to get killed.

Good risk: Gives you the best chance to survive and prosper over the long haul.

Good risk Bad risk
Simple to understand what the risk factors are, frequency and severity of bad outcomes, how they interact with the rest of your portfolio. Complicated, insufficient history to gauge frequency and severity of bad outcomes, possibility of a ‘catch.’
Good economics: properly compensated for the risk. Bad economics: poor reward for the risk profile.
Just because economics are good and you can get paid doesn’t mean you will get paid. Counterparties you can trust, who have limited ability to rip you off, and whose incentives are aligned with yours. Sketchy counterparties, with opportunities to change the terms of the deal, who have conflicts of interest, and who don’t care if you make money. Company managements can self-deal, sell out cheaply to a PE firm for rich management contracts. Financial counterparties can find fine print and fees to rip you off.
Risks asymmetrically skewed to the upside. Positive optionality/convexity. Limited downside, unlimited upside. Cheap long calls. Risks asymmetrically skewed to the downside. Negative optionality/convexity. Limited upside, unlimited downside. Cheap short puts. Bonds yielding 0%. Picking up pennies in front of a steamroller.
Naturally a hedge or diversifier – uncorrelated or negatively correlated with the rest of your assets under most scenarios. Positively correlated with your real liabilities. Texas hedge – positively correlated with your portfolio, negatively correlated with your liabilities.
No more risk than is commensurate with your edge, your ability to withstand losses, both financially and psychologically. Risk that exposes you to catastrophic blowup, or enough psychological pain that your judgment is impaired and you make bad decisions, don’t stick with your system, throw in the towel at the worst possible time.
Volatile short term, gives you the best chance of coming out ahead in the long run.1 Profitable short term, strong momentum, inevitably going to blow up at some point in the future.

The only reason you get paid more than T-bills in the stock market is because you are taking risk. How do you know for sure which side of the line you’re on?

You’re not going to know the difference your first day out. It takes time to get a feel for the financial and psychological toll the market can dish out, and what you can handle.

When you start, you need a system2 that limits the risk you take to what you are comfortable with. You need to do some math, either simple or complicated, that gives you an idea of the frequency and severity of bad outcomes or periods. And then you need to build experience. You need to cultivate the little voice that tells you, something has changed, those assumptions that you built into the system aren’t right for current conditions.

Risk means more things can happen than will happen – Elroy Dimson

Risk is ultimately subjective. You never really know all the things that could happen, or could have happened, and what the real odds are or were.

Just because you won doesn’t mean it was a good bet. Maybe you got lucky. Conversely, just because you got hurt doesn’t mean it was a bad bet — maybe you got unlucky.

The question is, was that the best risk-reward play? And in the long run, if you keep doing that, are you going to come out OK? Even a bet with positive expected value is a losing proposition in the long run if you bet too big. (The gambler’s curse.)

Regardless of how subjective risk is, poker players know who are the fish at the table, even if the dead money flops the nuts once or twice. Scouts know some athletic phenoms are not going to have a long career because they don’t have sound fundamentals or work ethic. Time and the law of large numbers and the central limit theorem convert the highly variable in the short run to the predictable in the long run.

George Soros claims he has developed a sixth sense for when something isn’t right with his positions, and he starts to feel physical back pain when he is not comfortable with his positions and tenses up. And yet he also famously said “it takes courage to be a pig.” When you’re right on something, you want to be be positioned to extract maximum value from being right.

Risk, pain, intense effort: instinctively most of us shun them. But your ability to face them with a healthy attitude determines your personal growth and success. We need to learn to appreciate the right kind of pain and risk and distinguish it from the wrong kind.

Risk is your friend when you’re getting paid the right price to take it, you put on the right amount in the context of your entire portfolio, lifestyle, expectations, and personality; and you monitor and manage it by diversifying and cutting when necessary.

Confidence is when you know what the worst case is and that you can handle it. When in doubt, get out, or limit your potential losses to what you can handle.

Volatility matters when you feel it. All the charts, ratios, and advanced math in the world mean nothing when you break down, vomit or cry due to the volatility in your portfolio. I call this the vomitility threshold.. Understanding your threshold is important, for it is at this point that you lose all confidence and throw in the towel.
– Ed Seykota

Man cannot remake himself without suffering, for he is both the marble and the sculptor. – Alexis Carrel

1 “You have never lost money in stocks over any 20-year period, but you have wiped out half your portfolio in bonds [after inflation]. So which is the riskier asset?” – Jeremy Siegel

2 The “system” doesn’t need to be complicated. It could be as simple as a robo-adivsor portfolio or a lazy portfolio. Or it could be a full-blown active trading system with criteria for market selection, position sizing, entries, and stops/exits. But regardless, you need to keep in mind what assumptions were made in picking the system and monitor that things haven’t changed in an important way, including your own life situation and risk tolerance, and market conditions which change the risk/return profile, such as a 1999 type tech bubble or current day ZIRP interest rates.

‘Net neutrality’, Netflix vs. the cable monopoly, and the Internet profits tax

Really, the way to understand ‘net neutrality’ is it’s all about Netflix.

The cable companies are outraged and scared to death about Netflix. If you’ve tried a Roku Internet TV appliance (or Apple TV, or Google Chromecast, or Amazon Fire TV), it’s a 10x user experience improvement on a cable box. For less money.

Netflix and cordcutting are hurting the cable TV bundle business model. Internet customers are growing, and TV customers are declining.

The idea that Internet TV could break the cable TV bundle and leave ISPs as a dumb Internet pipe is anathema to the cable companies.

The FCC made rules to prevent cable companies from blocking or throttling specific sites and services like Netflix. Verizon sued to overturn them. They won, the court said the FCC doesn’t have authority to impose rules like that, except under Title II, the phone regulatory framework, which hasn’t been applied to ISPs.

After winning in court, the cable companies throttled Netflix and made them pay for ‘peering.’

The argument that this has something to do with the costs that Netflix imposes is weak, very weak. Netflix is more than happy to build a data center next to Comcast, run a big pipe to Comcast, and pay for all their network equipment. That does not impact Netflix’s business model in the slightest.

And if Netflix customers use more bandwidth, the cable companies already charge the customers according to the speed and bandwidth they use, and if the costs are not in line, they can be adjusted accordingly.

If ISPs can charge any Internet service whatever they want, it will devolve to a tax on Internet service profits. They may or may not charge random Internet hosting services tiered rates based on speed. But by the time something grows to a Google or Amazon, they will have to negotiate one-off deals. And how much the cable companies can demand will depend on how profitable these services are.

And the key question you have to ask yourself is, if the cable companies could throttle Netflix or charge them to connect to their customers, could Netflix ever have gotten off the ground? And the answer, to me, is no chance. This is not hypothetical. ISPs have blocked apps that cost them money, like FaceTime, and wifi tethering. The tax the cable company would have to charge would have to reflect not just Netflix’s profits, but the customers the cable company loses.

So a world where Internet services have to get permission and pay to get in front of customers is not going to be the world of hot consumer Internet startups. (And who knows what happens to other services, if for instance, Verizon can charge companies to have employees work at home.)

So, the FCC, after a massive outpouring of consumer outrage, aided and abetted by John Oliver and the Internet industry, and an intervention from President Obama, is saying they will apply Title II.

Frankly, that’s (mostly) how free markets and democracy are supposed to work. What cable companies were proposing is an abuse of market power to restrain trade. What the FCC is doing is asserting its authority to maintain the status quo, after the cable companies pushed to tilt the playing field in their direction.

How is the over-the-top Internet TV world going to evolve? The $100-a-month bundle is going to be under pressure. I don’t watch sports, and I don’t want to pay $20 a month out of my bill for carriage fees for ESPN, YES, MSG, SNY, not to mention a bunch of other networks I don’t use. So I cut the cord about 5 years ago.

Is over-the-top going to be good for consumers? Bundling is complicated. There are good bundles and bad bundles. Microsoft can charge $100 for each of the 4 big products in MS Office. And people will buy 1.5 on average. Or price it at $200 for the bundle, which everybody prefers and makes more revenue for Microsoft. Or there is the music bundle, the LP album/CD. It turns out when people can buy singles for $1 a pop, revenue goes down.

Which do we think the cable TV bundle is closer to? Will people pay independently for Animal Planet or Shark Week? I kind of doubt it. I think you can safely short DISCA and SNI. Of course, it’s a hit-driven business, they could change format, have the next Mad Men, or sell out to an Al-Jazeera. The race is not always to the swift, nor the battle to the strong, but that’s the way to bet.

When the consumer decides, it’s pretty safe to assume they choose what’s good for them. And people will have the choice of skinny bundles or jam-packed bundles, cable TV bundles or over-the-top bundles or just a la carte individual services from HBO, CNBC, ESPN. And it will be good.

What about Netflix? I find myself watching more Amazon Prime than Netflix. Their bundle is that awesome. They have a strategic imperative to own digital media distribution, from books to music to video. They’re tremendous at execution. Netflix is rather fully priced at > 100x earnings. I think Netflix could get Amazoned, and it could be a long time before anyone makes any monopoly profits in this business, if ever.

PS. The talking points against applying Title II are breathtakingly cynical and self-serving. The FCC is applying 1930s telephone regulation in a naked power grab? So why did the industry sue against the lighter-touch regulation the FCC had in place before? Why did they force the FCC’s hand, so the FCC had to apply Title II just to maintain the status quo? It’s a problem that doesn’t exist? So why did ISPs throttle Netflix, why did telco ISPs block FaceTime, wifi tethering? Basically, Comcast and others say their position is, we’re for net neutrality, but Title II is the wrong solution. One one hand, you have Comcast saying, we’re not going to do anything bad, and you shouldn’t apply this broad regulation to us. And on the other the FCC is saying, we need to take this broad authority but we’re not going to do anything bad with it. Because it’s the only legal way you’ve left us to get you to do the things we’ve asked you to do in the past, like not blocking Facetime or throttling Netflix. Frankly, the FCC, and the millions who commented, are a lot more credible. Of course it’s about cable industry profits, and if you give them an incentive to do bad things, they will do them.

Andreessen v. Summers: Can you have robots, hoverboards, and secular stagnation?

Diane Coyle says you can have either robots, or secular stagnation, but not both. In a somewhat confused tweetstorm, Marc Andreessen says secular stagnation is BS. Larry Summers, who is one of the guys behind the secular stagnation hypothesis, responds. But then, confusingly, is reported to agree with Coyle.

While this is a statement one makes at one’s peril, I will say it anyway: Marc Andreessen is wrong, and it ties into his wrongness about Piketty.

Technology can be a very good complement to labor, or a very good substitute for labor.

The more a technology is human-like, the greater the elasticity of substitution between capital and labor.

In the extreme, consider a toy model economy where capital = human-like robots, and you can rent a human-like robot by the hour. Perfect substitution between capital and labor.

The wage rate is going to equalize with the hourly capital cost of the robot. If the cost of robots goes down, the robot rent and the wage rate both go down, all else equal.

Suppose the labor supply is fixed/perfectly inelastic. No departing the labor force when wages go down, no aging population, no population growth.

If you have a technology breakthrough and more/better robots for same price, then overall real labor income goes down.

So, as first order effects, when robots get better/cheaper, two things happen: there is more investment in capital, ie building more robots because they got cheaper. And labor income and consumption go down.

The question then becomes how much of each do you get, and what happens at the macro level?

Maybe after all the second- and nth-order effects work through, you have full employment. GDP increases due to increased supply of capital, but output shifts to investment, ie robots building more robots. In an extreme, you enter a singularity of faster GDP growth, wages going down, more and more robots get built to the point mostly you have robots building more robots, while consumption steadily declines, even as GDP rises.

Maybe you don’t have full employment. If animal spirits are not present and people don’t demand more robots because they don’t see sufficient end-user consumption demand, maybe there is an output gap, i.e. secular stagnation.

This is essentially the Piketty argument. If elasticity of substitution between capital and labor is greater than 1, labor gets relatively worse off over time as capital accumulates and technology improves, in a Solow growth model where capital and labor get paid their marginal product1.

It’s apparently hard to refute Piketty and show conclusively that the elasticity is < 1, either as a theoretical matter, or empirically. The best one can say is, historically it's been close to 1 in the very long run. Labor and capital shares haven't shown a consistent long-term trend either way. Historically, faced with technology that was a close substitute for labor, labor has ultimately done OK in the long run by specializing in what machines couldn't do (elasticity close to 1, very recent history notwithstanding). And historically, threats of technology making labor obsolete and specifically, how quickly artificial intelligence would improve, have proven to be over-hyped. Of course, until such time as we have fully autonomous android robots than can do everything humans can do, technology and capital are partly a substitute to labor, partly highly complementary, a force multiplier for labor. It would seem likely that over time the elasticity of substitution increases, as technology can more closely resemble human labor, perception, decision-making. You start with capital complementing and amplifying human labor, but as technology improves, it becomes more of a potential replacement. It seems impossible to conclusively refute that in the future elasticity is > 1, in the case of radically new technology that is a closer substitute for labor.

In the short run, surely even Andreessen would agree, more disruption means more structural unemployment. It’s the price we pay for productivity growth. Sure, a telegraph operator can retrain as a switchboard operator, and a good SABRE travel agent can retrain for other computer research, but it’s not good news for the travel agent/telegraph operator in the short run.

And in the long run, I think we’ll have to wait and see. Maybe we will find that capital is still a highly imperfect substitute for labor. Or maybe we will find that you can have hoverboards, self-driving cars2, and secular stagnation, and will have to figure out how to create jobs and distribute benefits of technological progress and growth.

P.S. As an aside, I find Summers’s faith in productivity statistics disturbing. In a time of disruption, productivity is hard to measure. A new BMW 3-series comes out. It’s the size of the old 5-series, has better mileage, side airbags, voice-controlled phone and navigation, traction/stability control, rear-facing video cam, heated seats, it lasts longer with less maintenance, I could go on. It costs more than the old 3-series. A ‘hedonic adjustment’ has to be applied. It’s not a conspiracy, someone has to make a judgment call, how much of the price change is inflation, how much is more car for the money. The BLS does the best job they know how, to say how much output went up, and how much price went up.

And that is a relatively easy industrial product to measure constant-dollar output. What if a new startup produces an electric, self-driving car? Deflate that.

And don’t get me started on measuring productivity in services. You are in a strange town, you have a hankering for Thai food, you fire up your phone, check menu and reviews, order Seamless. In the old days, you would have to find a dead-tree phone book, phone, talk to the restaurant, and take your chances. Suppose people switch from radio cars to Uber. You hit a button on your phone, car shows up minutes later, you pay less money for a better experience. Nothing is going to capture that productivity bump. Just fewer dispatchers and restaurant phone order-takers, which is not the real value-add.

Now, a line worker or secretary works for the car manufacturer. Her/his job is automated, robot assembly, no more phones to answer, copies to file, the engineers and lawyers do their own email and stuff gets filed in the cloud. So the manufacturer’s output of cars, however estimated, is divided by fewer people, productivity goes up.

Line worker gets new job managing a Cinnabon which is low output per hour. The economy’s overall productivity goes down, because the composition of output shifted to low-productivity services. The pressure on wages brings back a lot of services that didn’t even exist, when I was growing up upper-middle-class people didn’t have cleaning ladies, now they all do, and you can order all kinds of services on your phone. (Don’t get me started on Roomba output.)

Productivity sort of eats itself. Some are made more productive, others lose their jobs and get pushed into lower-productivity activities, erasing some of the benefit.

Or increased output doesn’t get measured at all. Auto-company paralegal does a project that involves 2 weeks of discovery in a warehouse. Technology turns it into a one-hour search. Maybe the company gets rid of paralegals and produces more cars per hour of labor. Just as likely, people do a lot more discovery. Does it make the cars any better or cheaper? No. Did the productivity evaporate into thin air? I don’t know. Is the economy better off? Depends on the value you place on that research. Maybe more better cases get made, more worse cases get defeated. Or maybe it’s a total waste. But the work and output is there, if not easily quantifiable.

Data only tells you so much.

I suspect there is some fundamental truth to the robots/globalization/inequality/secular stagnation nexus, but it will take decades to sort out and we’ll never really know for sure. You have to build the type of society you want and try to figure it out as you go along. There are always surprises and unintended consequences, and theory or ideology doesn’t reliably tell you what’s going to happen.

1 It’s interesting that Summers is arguing against Andreessen on the secular stagnation hypothesis, and against Piketty on r>g. To me, they seem to be two sides of the same coin. For good discussion of the whole Piketty debate, see:

2 Drivers have supplanted secretaries as the most common job in many states.

Game theory, Bill Belichick, Neville Chamberlain

There are some people that will be deterred by the fact that we have nuclear weapons… But those people are the folks we can deal with anyway. — General Charles Horner

How about that Super Bowl? Sometimes it pays to be irrational, to do the unexpected like pass on 2nd and 1, to catch the defense by surprise and force them to defend the pass. By the numbers, Carroll should have been running out the clock, and Belichick should have been calling timeout to give Brady a chance for a long pass and field goal, if Seahawks scored quickly. But Belichick says he felt running time down to where Seahawks had to call a pass was the way to go. And when the Seahawks called it, the Patriots were prepared. On paper, the pass isn’t a terrible call if it keeps the opponent guessing, and you don’t have time for 3 running plays. But if one believes the evil genius of Belichick, he psyched Carroll into calling it, and it didn’t surprise anyone.

  • Game theory only works if you’re dealing with rational people. Not with dumb, ideological, or crazy people.
  • Most people are only rational about unimportant things. On the things that matter most, they’re usually emotional, ideological, stupid or crazy.
  • Therefore, game theory is only useful in dealing with unimportant things.
  • By being irrational, you get your opponent to throw out part of the toolkit, and have to consider and defend a lot of otherwise illogical actions. So ironically, in game theory it can be rational to be irrational. If you’re on a one-lane road and you want everyone else to get out of your way, slobbering at the mouth or just throwing the steering wheel out the window will do the trick nicely.
  • Which leads to the problem that you never know if your adversary is pretending to be crazy to get his way, or really is crazy.
  • On average, it’s more sensible and profitable to assume that the adversary is rational.
  • If you assume that your wartime adversary is insane, then really the only possible outcomes are 1) caving to their insanity or 2) their total destruction (or yours).
  • Always assuming their insanity is tactical rather than congenital therefore yields better results, and has the benefit of discouraging everyone from crazy behavior, since it isn’t taken too seriously.
  • Of course, every so often you run into someone who really is crazy, e.g. Hitler. And history hasn’t been kind to Neville Chamberlain, who people regard as a cowardly appeaser, when in fact he was a cold-eyed Conservative ‘realist’. (History can be so complicated… Edward VIII was pro-Nazi (along with Henry Ford and Joe Kennedy)…and George VI, if not pro-Nazi, gave Chamberlain an extraordinary photo-op and political endorsement by whisking him from the airport to Buckingham Palace to wave and prattle about ‘peace in our time.’)
  • We’re better off living in a world of rational people, who assume others are rational. Perhaps, giving the occasional Hitler a little too much leeway is the price to be paid for living in an world where most people act rationally most of the time and expect others to do so.
  • I certainly understand, if people whose ancestors were at Auschwitz don’t agree with that. But I wouldn’t run my foreign policy on what they think, or for that matter on what any other foreign power with their own interests happens to think. When you live like everyone is irrationally out to get you, you create a reality where a lot of people are quite rationally out to get you.

A Greece reading list (Or why the euro is doomed)

Time converts the improbable to the inevitable – Stephen Jay Gould

[TL;DR 50% odds Greece leaves euro this year. Odds eurozone breaks up eventually: 100%]

If you don’t care too much about the Super Bowl today, here are some things you could be reading about Greece:

To summarize:

1) Greece cannot service its debt.

Admittedly, the rate Greece actually pays is much lower, but you get the picture. Even at 2% rates, in a zero-inflation environment, Greece would have a tough time.

2) What cannot be repaid will not be repaid (Martin Wolf). So why does Europe insist on no debt reduction? Two reasons:

  • EU politics: Debt is a cudgel to exert political control over Greece. Every so often, Greece has to come back to Europe for a round of ‘extend and pretend’, which Europe hinges on political ‘reform’ conditions.
  • Domestic politics: Merkel and Eurozone leaders don’t want their political opponents to claim voters’ tax dollars are bailing out Greece.

3) The ‘profligate Greeks’ is only very partly true.

Tax rates in Greece are comparable with the rest of Europe. Greek workers work the longest hours in Europe.

The main issue is, the upper classes and the oligarchs don’t pay their taxes. (Death threats forced me to quit: Greek tax head.) That’s why debt went to 100% of GDP. It went to 175% as GDP shrank 25% and the debt was rolled over in the bailouts, which were really bailouts of European banks that would have gone broke if Greece defaulted.

So, if you’re an oligarchic shipping magnate, when Greece went into the euro at a too-high drachma rate, you were able to offshore your fortune in a hard currency at a great rate, while benefiting from living in a corrupt tax haven.

If you were a low income worker, then tourism, agriculture like yogurt and olive oil took a hit from moving into a stronger currency. You still paid your 20% VAT. Then when the economy went south, you lost your job, with unemployment reaching 28%, people burning garbage to stay warm, global pharma companies halting shipment of drugs to hospitals because they weren’t getting paid.

Far from profligacy, ordinary Greeks have paid an extraordinary price. Being stuck in the euro, Greece didn’t get benefits of either default or devaluation that would help tourism, agriculture, manufacturing. Benefits of the bailout went to euro banks, Greek oligarchs.

5) End game

From Greece’s standpoint:

  • Status quo is unacceptable: 25% unemployment, inability to pay debt, perpetual harsh bailout conditions.
  • Default and euro exit would be another disaster with unpredictable consequences.
  • Greece would be a financial pariah state.
  • Potential for very high inflation in a return to the drachma without access to global finance.
  • Unpredictable and potentially very high cost in the short run, but default and weakening of currency would bring back tourism, agriculture, export industry, and pave the way to recovery in the medium term.

From Germany’s standpoint:

  • Hard to offer debt forgiveness, less austerity to Greece while demanding same from rest of periphery.
  • Domestic politics of taxpayer money going to Greeks.
  • A euro exit would be very damaging to the euro project. Possibly fatal in the long run, as bank deposits in Italy, Portugal and Spain would be viewed as less safe than bank deposits in Germany. In effect, southern euros would not be the same as German euros.
  • Geopolitical factors figure strongly as well. No one wants Russia or China to establish a foothold in the heart of the Mediterranean. (Russia’s navy access to the Mediterranean is a factor behind the messes in Syria and Ukraine/Crimea.)

Clearly there is a deal to be made. Forgive unpayable debt over time. Allow a little more fiscal space to ease the burden on ordinary Greeks. Continue structural reform focused on bringing oligarchs and affluent into tax system, privatizing without fire sales.

It’s an ultimatum game.

6) How does the ultimatum game get resolved?

I don’t know. 50/50 a deal, or one party defects, or events overtake both of them, like a run on banks, bank holiday, and market collapse.

It all hinges on how each party feels about how bad the best alternative to a negotiated agreement (BATNA) would be, how much the parties trust each other to uphold an agreement (Does Syriza have the will and competence to accomplish structural reform?), how credibly they are able to communicate what the true red line is.

The clock is ticking. Every time Tsipras draws a line in the sand, more euros will flee Greek banks to Germany, forcing ECB to replace them with liquidity assistance which would, of course, go up in smoke in the event of a default.

At some point, Tsipras’s best deal is Grexit, and blame Germany. Likewise, at some point it’s better for Merkel to say Grexit was the Greeks’ fault, they’re an exceptional case and this could never happen to Spain, Italy and Portugal , and take steps to ring-fence the periphery and their banks.

7) Will the euro hold together in the long run?

Even if Greece stays in the euro, it’s hard to ignore that they have some developing-world political and economic dynamics and only heroic measures will have kept them in the euro.

Ultimately, the euro only works in the long run in the context of full political, fiscal and economic European integration.

It’s worth noting that QE only came to pass as a result of Draghi using all his political capital, including possibly floating resignation threats. Draghi himself has said the eurozone does not meet the minimum criteria for sustainability, economic and fiscal union are needed.

It’s as if the US entered into a currency union with Mexico.

The US and Mexico have divergent economies, but separate currencies. Suppose both countries make cars. Mexico is, hypothetically, a poorly performing economy, with poorly educated, unproductive workers, inept management, high taxes, inefficient government regulation and services, resulting in low quality, expensive cars.

Mexico will have a hard time selling its cars domestically and exporting them to the US. The US will export cars to Mexico.

As consumers buy fewer Mexican cars and investors have little reason to invest in car factories there, the Mexican dollar will tend to decline vs. the US dollar. But as the peso declines, Mexican cars will start to get cheaper relative to US cars in both countries, allowing more to be sold. Poor productivity growth leaves Mexicans with a lower relative standard of living as imports get expensive, but they can still sell cars and maintain employment, and the adjustment in the currency cushions the impact on production and the domestic economy.

Now suppose the two places we’re talking about are Michigan and Alabama, which share a currency and a common Federal government. Alabama’s car manufacturers do well, Michigan’s do poorly. Michigan can’t devalue its currency, since it shares the dollar with Alabama. Its auto sector shrinks. People are laid off from Michigan car factories, and some of them move to Alabama, where car workers are in demand. The overall US economy is doing no better or worse than before, and as Alabama booms, taxes are generated, and help the Federal government pay for unemployment benefits, retraining, pension benefit guarantees, social security as older workers retire, and other help for Michigan. There is no currency cushion, but labor mobility, sharing of taxes and transfer payments cushion the adjustment.

Now let’s return to the first Mexican example, and suppose the US and Mexico shared a currency, the ‘amero’. Mexico can’t devalue, so its cars remain uncompetitive and manufacturing shrinks. Workers lose jobs, but can’t easily move to the US due to language and cultural barriers. Mexico’s unemployment rate goes up, tax receipts decline and the Mexicans have to borrow to pay unemployment benefits and cut services. They start to run deficits and eventually they’re on the brink of default and ‘amero’ exit. They turn to the US for help. US taxpayers feel they did everything right, so why should they bail out those lazy Mexicans? Mexican taxpayers feel like those hegemonic Americans are telling them what to do, and running a currency and interest rate policy that impoverishes Mexicans.

When the ‘amero-zone’ status quo becomes politically unsustainable, one of two things can happen. 1) The Mexicans go back to the peso, which depreciates sharply, making it impossible to pay back dollar debts. They default on sovereign debt. They institute capital controls which make it impossible for their companies to pay back ‘ameros’, and maybe rewrite the rules so foreign debtholders can’t take the assets in bankruptcy.

Or 2) the US goes back on the dollar and Mexico stays on the amero, which depreciates sharply. Now the Mexicans can pay back their debts in cheap ameros, and there is no need for a messy default.

Either way, any poor suckers who end up with Mexican assets and US debts get taken to the cleaners, including banks which will be busted and have to be taken over by the government or get massive bailouts.

Substitute EUR for amero, Greek drachmas (and Irish punts and Portuguese escudos) for pesos, deutschemarks for dollars, and you get the picture.

The fundamental problem is that, if Greece shares the euro, and it doesn’t have an FX policy or monetary policy, there needs to be some mechanism to drive economic convergence. If not, it will have the wrong policy at the wrong time: 28% unemployment and a pro-cyclical policy of more austerity the worse things get.

Europe’s monetary integration is way too far ahead of the economic, social and political integration. In fact, I just don’t think Europeans want American-style political and economic integration. Perhaps a bailout can be arranged this time, but the next one may be politically and economically impossible.

Unless the political and economic integration catches up, a sufficiently big crisis will inevitably rip up the Euro zone. Either a hard core centered in Germany will secede from the Euro, which seems less damaging and therefore more likely, or some peripheral countries will undergo Latin American style political and economic upheavals (or both).

Maybe we’re witnessing the slow motion breakup of the euro zone. Or maybe the Big One is still out there.

Either way, it will be a big mess.

PCs Were the Triumph of the Nerds; iPhone is the Revenge of the Cool Kids

Apple reported a blowout iPhone 6 launch quarter. In fact, reportedly the largest quarterly profit ever reported by a public company. I had a feeling they would blow away expectations, the iPhone 6 looks and feels great, it’s a must-have upgrade.
download

Apple now has, contra Steve Jobs, an iPhone/iPad for every size and pocketbook. Not to mention deep inroads in fast-growing economies where it’s very bling-y. Everyone I know (mostly rich white NYC folks) either has one or wants one.

And I’m an Android guy, cheap and tech-savvy, and I want one too!

The current crop of flagship Android devices (Samsung Galaxy S5, Moto X, HTC One, OnePlus One) has better specs for less money than iPhone. But they lack excitement. They’re S3’s with better specs. I want to try Apple Pay and see if it’s the future, even the Apple Watch.

I’ll probably wait for the Samsung Galaxy S6 in the spring. The specs are nuts and the wireless charging is cool. I really like being able to take out the SD card and battery. And I really like the idea of a phone where if I lose it, I can probably replace it pretty cheap on Ebay.

But I feel the irresistible pull of the iPhone.

The biggest thing the iPhone has going for it is, it’s the phone for the cool kids.

All those folks (mostly rich white NYC folks) that, for professional and social reasons I would like to impress with my brilliance and tech savvy and general with-it-ness, are iPhone now, just like they once were BlackBerry. And Android and Samsung are sort of the poor cousins.
Zoolander

And that’s the enviable market position of Apple: It’s the revenge of the cool kids.

Microsoft’s triumph was driven by standards and economics. Corporate IT picked IBM and hence Microsoft. Clone competition drove costs down. Scale and platform effects made PCs and eventually Windows ubiquitous.

The iPhone is driven by the consumers. No IT folks pushed iPhones. End users demanded it.

The iPhone has a healthy app ecosystem. Android…well… all the cool apps come out on iOS first and Android is an afterthought, unless there’s some Google juice involved.

Why? For one thing, the cool kids include the early adopters with disposable income, and that’s who app developers target, and they buy iPhones. For another, Android development is hard, a pain, a mess. It’s a more fragmented platform, more OS versions, form factors, hardware configs in the wild to deal with. And the toolchain sucks.

AndroidCaptureCapture

Android matches iPhone feature for feature. A lot of iPhone features (voice input/control, notifications, widgets, big screens) came from Android. iPhone may be a slightly slicker user experience. But iPhone has one ace in the hole: A better app ecosystem, and the next Instagram/Snapchat/Yo will not show up first on Android.

This drives a virtuous “God device” cycle for the iPhone.

1. ‘Cool kids’ buy it
→ 2. Apps target ‘cool kids’ who have money, who are early adopters, who drive adoption of Instagram/Snapchat/Yo
→ 3. iPhone is seen as the ‘cool kid’ phone. It’s a brand/fashion dynamic: Hermès v. Jos. A. Bank, Rolex v. Timex, Macallan v. Budweiser.
→ 6. Apple can charge a premium for it, it generates stupendous margins.
→ 7. Apple reinvests and drives its unitary platform forward to payments, watches, while Google has to corral hardware manufacturers, carriers, can’t worry about friggin’ banks.
→ 8. iPhone is even more cool. Go to 1.

At one time I thought: Microsoft is the new IBM. Google is the new Microsoft. Apple is, well, the new Apple. Makes highly differentiated products at a premium, beloved by cognoscenti, too expensive, too loathed by OEMs (phone companies) to own the smartphone market when it’s mainstream.

But instead, over time, the Android and iPhone user bases diverge more and more, which diverges the branding and to a lesser extent the app ecosystems. Fragmentation and competition make Android cheap. But decreasingly relevant, as they can’t drive apps like payments mainstream, or categories like watches, maybe virtual reality.

And that’s why Apple earns more net income than Microsoft and Google combined, and on net the rest of the mobile phone industry makes no money.

Is it a defensible position for Apple? Well, the insane margins have been declining, but Apple has been more than making it up on volume growth, even as its share of smartphone units is pretty flat.

I’m not saying it will go on forever. But it seems clear the margins won’t get competed away to nothing, like in the rest of the market. At least until something better blows iPhone out of the water.

And there are opportunities for growth. Payments could be huge. Of course in the future you’ll pay with your phone, and you’ll get offers, manage loyalty programs on your phone.

And who knows if something like watches could move the needle of incremental growth, or be a monster, the next big thing.

But we’re in an era no longer dominated by new technologies opening new vistas. Instead, it’s about making insanely great products with the technology platform we have, reinventing communications (WhatsApp, Snapchat, Instagram) and industries from the ground up (Uber) and driving adoption with a viral brand/fashion dynamic. And that tends to play into Apple’s hands.

The Dark Web Stack, Or How To Eff Up The Net

You have a choice of trusting the natural stability of gold or the honesty and intelligence of members of the government, and with all due respect to these gentleman, I advise you as long as the capitalist system lasts, vote for gold. – George Bernard Shaw

Do You Want To Eff Up The Net? Because That's How You Eff Up The NetAn interesting dive into “Deep Web Marketplaces” by the folks at avc.com and USV.

A deep web marketplace is like the Ebay of anonymous e-commerce. Other elements of the stack include:

  • Payments: Bitcoin
  • Transactional database: the blockchain
  • Non-transactional text database1: Pastebin (Images: Imgur; generic NoSQL database… I don’t know of one, so build on top of those, or left an exercise for an aspiring entrepreneur)
  • Networking: TOR
  • CDN: Bittorrent, Pirate Bay

Could you could go down the list of Amazon Web Services and come up with p2p distributed versions of each one? I don’t see why not.

Could you have a distributed version of Heroku/Amazon Web Services, an anonymous, distributed platform with all the services and APIs to create any app or marketplace from Ebay to Uber running in your browser or on your phone? I don’t see why not.

Could you have a distributed p2p version of UPS like TOR, with people handing each other anonymous packages and delivering in some dead drop or to the holder of dollar bill number B12345678?

Of course, that’s what they would do in an underworld network, or in a totalitarian state. Like samizdat publishing.

Should we?

There’s a constant ebb and flow between centralization (mainframes/AOL) and decentralization (PCs/Web) and back (Mobile/cloud).

Similarly there has been an ebb and flow from relatively anonymous Web protocols like Web, SMTP etc. to trusted IDs, Twitter/Facebook/SSH, and back with Bitcoin and dark web.

There’s a dichotomy: A world where much of the communication, transactions, commerce have to be over a dark web would be a pretty effed up place, like one where people had to pass along literature through samizdat and do commerce in back alleys.

The dark web stack is kind of an effed version of the legit web, and yet necessary.

It’s just that, being nexus-free, it’s more resistant to control.

Not totally resistant…the Internet is the granddaddy of distributed, anonymous networks, and it’s impressive what China can do to control it.

What can a dark web stack do, that a more centralized client/server or mobile/cloud stack can’t?

Absolutely nothing.

How much of your life needs to be conducted in a secure, ‘erasable’ Internet? Sure, some part of communications should be kept very private, and quickly forgotten.

Some, but by no means most of it.

And there’s a cost. It may avoid an ATM charge, but the cost of clearing a Bitcoin transaction is still staggering compared to institutional FX markets. The time to clear is staggering. And fairly fixing any glitch in centralized markets is maybe tough and possibly political, but it’s practically impossible on the blockchain.

And yet, it’s insane for someone like David Cameron to say the government needs the keys to everything and there can be no true dark web. If you had to put a back door in every communication or ecommerce system, impossible to believe anything would be safe against black hat hackers and foreign governments. And of course bad guys would always find a way around it. And there’s a free speech issue: what kind of world is it where you can’t have a private conversation in the safety of your own device? It’s unachievable, dangerous to try, and wrong.

The more you use control of legit platforms to enforce political goals, whether it’s against Falun Gong or Russia, the more you create demand for a dark, private web.

If people trust the government and the monetary and financial system, they don’t hoard gold. If they hoard gold, the response is to fix the monetary system, not to ban the ownership of gold. Which may be the only alternative when the monetary system is irretrievably broken in times of war or crisis. But it’s not the way a free economy and democracy are supposed to work. And the same goes for the dark web.

If you don’t want people to use the dark net, don’t mess up the legit networks with back doors and warrantless wiretaps, ‘express lanes,’ censorship, using them for political pressure. That’s not the way freedom of speech and democracy are supposed to work.

Or people will create worse versions and route around you.


1 Technically, these aren’t distributed in the same sense as the blockchain is. Many distributed apps and use cases could probably use them as a storage layer, though. For instance, to build a distributed p2p Uber, drivers could have an app that posts availability and reservation responses to a Pastebin type public space signed with their key, and riders could likewise post reservation requests. Perhaps there’s an opportunity for a NoSQL nontransactional distributed p2p database counterpart to the blockchain, which is transactional, strictly enforces no double-spend, but takes a long time to commit.


26 queries in 0.216 seconds.