Mauboussin on Success and Luck

My 5-year old daughter recently was very excited to discover a new solution to a pressing problem. Her mother had told her she was not allowed to eat cookies and left the room. 5 minutes later Izzie told me she had a great idea: I give her a cookie but we don't tell mom, so mom wouldn't be upset.  Win-win! Her pride in independently discovering the tactic of lying reminded me that many skills are no less interesting to others simply because I know them.  Such it is with Michael Mauboussin's book, The Success Equation, which investigates how to navigate in a world filled with skill and chance.

The book spends a couple pages explaining the reversion to the mean by noting Galton's original observation on the heights of parents and their children, and much of the book involves the concept of bayesian updating. These aren't new ideas.  Yet, when he noted that while small schools are overrepresented in data on excellent and really bad schools, it reminded me that this is a real problem because invariably they are used as templates for some system-wide initiative, neglecting the fact that often some very unique circumstances as opposed to any efficient method are at work. Like everyone, I need reminding of some of these patterns sometimes.

The author recounts a fortunate interview he had with Drexel back in the days this was one of the most coveted positions out of college. He astutely noticed the senior interviewer had a Redskins trash can and so mentioned something casually about football, after which the interview went extremely well, talking mainly about football.  Clearly, that wasn't just luck, but after the fact appeared the key to him getting hired, creating a career path that would never be the same. Most big events, good and bad, are a mix of skill and chance.

In dealing with success and luck Mauboussin presents a useful model:

estimate=population mean + c*(sample mean - population mean)

c is a constant between 0 and 1, 1 for activities involving all skill, 0 for those that are purely chance. So, a chess, which is mainly skill, should have a c, or 'shrinkage factor', near 1, hockey games something near 0.  This is like updating a prior belief with new data (see here for how to do this more formally for various distributions). More relevant to investing, if your backtest generates a Sharpe of 2, one should remember most investment strategies have Sharpe's near 0, and so adjust your sample mean (ie, backtest) closer to that population mean. Any new idea does worse than its backtests because these aren't population data, but rather selective sample data.

He could have added something on how to address robustness as he did with his shrinkage factor approach to bayesian updating, say by showing how you could optimize over a set of parameters from different subsamples.

The book mentions research by Pinker, Gazzaniga, and Haidt, and I love these authors so I found all that very interesting.  Some of his other inspirations I'm less fond about. He calls Peter Bernstein 'one of the investment industry's greatest thinkers', though I'm at a loss as to what Bernstein insight might deserve such praise.  And then he mentions my favorite flâneur, Nassim Taleb, and suggests his great insight is that when payoffs are uncertain and complex, these are great situations to go long, specifically, go long out-of-the-money options.

Out-of-the-money options aren't underpriced on average. Sure, you will do well when the market tanks, but it's expensive insurance. Remember, when you cross the spread on a 3-delta put, your are giving away a lot of vig. If you think you can trade at mid or close you are being naive (I hear Taleb has a new book out advocating the same thing, buying gamma).

Also riffing on Taleb, he says we are better off using no model than a faulty one. This is a straw man. A faulty model is harmful almost by definition.  Yet  if you are acting, you are acting on some kind of theory, which is at some level a kind of model. One is better served by an attempt to formalize as well as possible one's strategy rather than to merely trade on intuition, the key being how well one deals with the uncertainty by simplifying the functional form, or accurately calibrating the probability of loss, or the time series nature of losses, etc. When people say they have an atheoretical or non-parametric approach they are neglecting a meta-assumption: any idea about how the world behaves involves a theory at some level, which can also be some kind of model (eg, a Venn diagram, or an if-then statement, a linear best-fit through an ellipsoid of data).

I'm not the audience for this book, but as most people are prone to overfitting, it has an audience, such as people who like Taleb's books.

Lastly, I noted he has 5 children, and I'm always  amazed when I read a sustained argument written by someone around my age with more kids than me. Having children is like putting a bowling alley in your head, and makes writing very difficult.  That's skill.

Missing Risk Premium Book Reviews

My book has 8 reviews over on Amazon, including the prolific Aaron Brown and David Merkel.  There's also some guy named Gregory Fodor who gave it 5 stars, the same score he gave to Ilmanen's Expected Returns, and a book on UFOs (I do find Ancient Alien Astronauts very fun).

I've noticed several others across the web. Here are some snippets.

Kirkus reviews:

The author’s conclusion is likely to be controversial in some circles, if not downright inflammatory

 John E. Parsons:

Questioning the theory is a tough challenge, though. As already mentioned, the risk premium is a central premise, and modern finance theory is a ramified structure. If we remove that premise, there is a lot of work to be done to recreate the structure around an alternative. I was surprised to find that Falkenstein understands the burden his questioning entails, and this book is a partial attempt to flesh out an alternative. Falkenstein is a serious fellow, and he has engaged the problem persistently over many years, so it is interesting to listen to his suggestions.

In this book, he provides a lucid history of the academic thinking on risk and return over more than a half century, a careful exposition of what the data say and don’t say, and a thoughtful discussion of competing theoretical frameworks. He writes with the refreshing voice of an outsider, and looks upon academia with a gimlet eye. But he also writes with the qualifications of an insider, completely familiar with the most sophisticated economic theories and statistical tools. He assumes the burden of making his critique and ideas resonant to an open minded intellectual familiar with modern economics. Make no mistake, this is a wonkish book that places serious demands on its readers.

Eddy Elfbein:

One of the most eye-raising aspects of the book is where Falkenstein discusses the many small losses that individual investors suffer between the stock gains they see reported on CNBC, and the returns they get. Investors are constantly dinged by things like bad timing, transactional costs, bid-ask spreads and taxes. Once you throw in variables like survivorship bias, Falkenstein says that the historical databases we have return bare little resemblance to what made its way towards investors’ pockets. This topic alone could serve as a useful book.

OnlyVix

The book is mostly technical, but without unnecessary math, and is focused on the main thesis - there is no "investment edge" in simply taking the risk. I would recommend every investor keep this in mind.

Value and Opportunity:

Summary:
+ the book is a good summary of all the current available studies which contradict the CAPM
+ he makes a good case for investing in low volatility assets, although I didn’t fully understand his theory
- what he misses in my opinion is the fact, that all this is common knowledge among value investors.

Blaming the Victim

Every week brings a new billion dollar settlement against the banks that supposedly rammed this stuff down the throats of the noble savages who couldn't afford the homes they bought.  Luigi Zingales notes a study on how culpable banks were:

Predatory lending is commonly defined as lending that imposes unfair and abusive loan terms on borrowers. The immediate effect of this mandatory counseling was to discourage almost half of the loans. ... While the paper calls it predatory lending, the better term would be crazy lending, because it is not clear that the banks were forcing these loans on ignorant consumers, and not the other way around. In fact, the reason why the program was suspended only 20 weeks after its beginning was that the local population complained with the legislators for the negative effects this law had on the availability of mortgages.

I wonder what happened to all those community organizers aligned with ACORN who advocated more no-down payment loans?

The head of CountryWide continually bragged about how they were easing mortgage requirements in the name of increasing minority homeownership when he was doing it. So, their public stance went from being admirable to brazen. This same government that applauds excess and piles on when things go wrong is supposed to moderate financial cycles.

AQR on Making Low Vol Better

 AQR has been offering 'low vol' funds institutionally for a couple of years, but last July started offering them to retail customers too.  As they are titled 'Defensive Equity' funds, they escaped my radar but really these are 'low vol' funds (eg, U.S. Defensive Equity Fund AUEIX, the AQR International Defensive Equity Fund (ANDIX), and the AQR Emerging Defensive Equity Fund (AZEIX)).

I spoke with Adrea Frazzini (right), a portfolio manager at AQR, and one of the lead authors on a major bit of research in this area, Betting Against Beta.  He noted their approach is basically to minimize volatility while trying to retain industry and factor neutrality, where the factors are such things as value, size and momentum. Considering that many think low vol is redundant in the context of a value exposure, this approach would appeal to those who already have a value exposue.

The idea here is that this approach generates less  tracking error that would come from, say, the SPLV, the popular low vol ETF that just swipes the bottom 100 stocks from the SP500. Basically, SPLV incidentally takes on various amounts of factor risk at any one time as some industries or extreme-factor loading stocks become less volatile.  By minimizing factor exposures tracking error will be less correlated with factors, which I can empathize with because if you underperform by 5% one year it feels better if its from something you can't control (the serenity prayer).

There's great danger and opportunity on this path. That is, SPLV is incredibly simple but it gets you a long way towards capturing the low vol effect: much lower vol, perhaps a slight return bump. In contrast, Russell's failed LVOL ETF tried to take it to the next level by getting really complicated, although I can see how, with a bunch of PhDs in a room it all made perfect sense. SPLV nicely keeps US low vol funds focused.

Like the other strategies in the low volatility space AQR expects betas of around 0.7, and volatility of about 1/3 lower than their regional equity benchmarks. Such a claim is realistic because that has been the result of other funds that have plied this strategy in real time over the past 7 years. As long as the Security Market Line (Beta on x-axis, average return on y-axis) is empirically flat this approach seems a no-brainer to any MBA because you can get the same return for 70% of the volatility; a low beta/vol focus has a higher Sharpe. [I'd say things about AQR's performance, but our far-sighted regulators make that parlous in their efforts to protect widows and orphans].

Frazzini mentioned that in non-developed countries, the volatility reduction is even greater within those regions, and clearly for any investor taking advantage of international diversification seems a straightforward way to reduce volatility and covariance with one's income much more than staying local.

Wittiest Sentence I've Read This Year

Steve Pinker attempts to explain the difference between Red and Blue states, and also why gays, guns, and taxes are correlated policy positions. In the process he notes that conservatives tend to have a pessimistic vision of human nature, liberals, a more optimistic notion:

The metaphors may be corollaries of the tragic and utopian visions, since different parenting practices are called for depending on whether you think of children as noble savages or as nasty, brutish and short.

Buy on Earnings Guidance

The standard finding is that a stock’s cumulative abnormal returns drifts in the direction of an earnings surprise for several weeks following an earnings announcement. Often this was presented as a way to make easy money shorting the negative surprises, going long the positive surprises. Now it seems one leg might have the wrong sign. This highlights one problem with trying to be consistent with all the facts: they aren't all true.

There's a newly published paper by Das, Kim and Patro On the Anomalous Stock Price Response to Management Earnings Forecasts. From the abstract:

In the post-announcement period, we find a significant upward price drift for both good news forecasts and bad news forecasts.

Here's the result in a nutshell, x=0 is time of earnings announcement,  the blue line negative surprises, red positive, going out 30 days

A free early version is here.

Arbitraging Beta Within Mutual Funds

If you remember your Corporate Finance, returns should rise (on average) with beta, because otherwise you can form a levered position generating a greater return without more risk. Arbitrage Pricing Theory developed by Stephen Ross was based on the idea arbitrage and totally consistent with the Capital Asset Pricing Model because it was simply the special case where one is arbitraging a single 'market' factor. A recent SSRN paper, Capitalizing on the Greatest Anomaly in Finance with Mutual Funds by David Nanigian.  Here's the Average return by beta quintile for US mutual funds, 1990-2012 (Figure 2--I added the red to see his points better:

As the author points out, there's an arbitrage here in Sharpe space, because you can generate the same beta and/or volatility by levering up a low beta mutual fund, and then generate another 2.5% return here. But, the best way to play this is less indirect via low volatility funds, because most players in that space have generated an even larger arbitrage opportunity in risk-return space. The key is beta is clearly not priced within equity markets, but it should be, so in the meantime anyone who doesn't have a really good reason to believe they have alpha should invest in a low volatility fund

Pointless Econ Debates

A big point of contentions seems to be whether or not the 2008 slump was a 'financial-induced recession', because as Reinhart and Rogoff say these have slower-than-average recoveries, this benchmark then makes the current recovery look better; in contrast, compared to the average recession the recent recovery looks worse.

 My old mentor Hyman Minsky argued that most recessions are financially induced, from excessive leverage, creating a panic as investors stop rolling over debt and cash flow long-since became negative. The Ponzi borrower borrows based on the belief that the appreciation of the value of the asset will be sufficient to refinance the debt but could not make sufficient payments on interest or principal with the cash flow from investments; only the appreciating asset value can keep the Ponzi borrower afloat. This applies perfectly to many zero-down home buyers circa 2005-7.

 Reading Minsky you'll see he applied this mechanism to most recessions (eg, not 1945), and so there's at least one expert--of good faith, not ignorant--who disagrees with Reinhart and Rogoff. Anyway, John Taylor classifies 1882, 1893, 1907, 1913, 1929, 1973, 1981 and 1990 as 'financially induced.' Krugman says 73 and 81 were not 'financially induced' because they were orchestrated to combat inflation, a novel addition to the classification. Then they get into where to start the recovery, and how far out to go. When the debate gets into such semantics, it becomes pointless.

 Krugman laments that 'politicization is hurting economics', though he is economist number 1 for having prestige and throwing around ad hominem like liar and consistently assuming that those he disagrees with cannot possibly be intelligent and honest. DeLong's posting are even better for over-the-top ad hominem, but he's in such a strange echo chamber he doesn't realize he's disqualified himself from so many positions he covets because even his team understands that rank partisanship becomes a liability at some point (no more Assistant Deputy Secretary positions, but rather, senior adviser posts for groups like the Rent is Too Damn High Party).

You just have to stop and remember: on any big debate, to think that one side is only motivated by ignorance or deceit doesn't understand the debate. Sure, some, even many, on any side of a big debate are ignorant are tendentious, but as with benchmarking this past recession, reasonable people can quibble with vague categorical definitions.

 I enjoy reading the daily venom from Krugman and DeLong because they are proud to be angry as if their righteous indignation makes them more compelling, unaware that losing one's temper is a good signal you've lost the debate. 

Merton vs. Low Vol

Low volatility investing is becoming more popular, but the question is perhaps it could be better captured via a more inclusive metric of volatility. The Merton model of default popularized by Moody's KMV is basically a function of two inputs: volatility and leverage. If this model is correct, then a probability of firm failure is better captured than mere volatility alone, and perhaps it also captures the true, fundamental volatility that is driving the low vol effect. 

I took data on the distance-to-default (DD) that I calculated monthly using the standard approach (I used to work on default models at Moody's), and this isn't really ambiguous: anyone in the know can generate them. I then looked at the top 1500 stocks by market cap, and formed portfolios with the highest volatility/lowest distance-to-default, and those with the lowest. Remember that low distance to default means 'risky', high DD is low risk. It turns out the portfolio returns generated from this exercise were so similar, the low volatility/high distance to default portfolio lines are indistinguishable.

It seems for the purpose of predicting future equity returns volatility and the Merton model are synonymous.  

Google Stop Per Regulations

Google had a PR snafu and announced earnings intraday. This caused a big price move of 10%, and then a price freeze for about 2 hours per US regulations.



To think this had any help on generating a more efficient market is absurd. This disruption wasn't horrible but highlights the futility of trying to do good: such regulations always ends up something really dumb that just annoys everyone in the know.  

I Would Destroy Half the Value for Half the Price

Above is the stock price over Vikram Pandit's CEO tenure. He received $1 in 2009 and 2010, but then you knew he would make that up and so in March 2011 he was given a $23MM 'retention award'. As CEO, he has a lot of power so it's hard to avoid this, and to me it highlights the problems of allocating incentives and rights in large collectives. While I'm a critic of these poor CEOs getting large payouts, I don't think the solution is obvious. After all, Clinton's millionaire tax in circa 1994 gave birth to the stock option boom that probably exacerbated the tech bubble and the fraud in companies like WorldCom and Enron.

A lot of people find Pandit's pay extremely annoying as if Pandit is taking their money, but unlike government money, this is other people wasting their own money. Sure, Citi was backstopped in the bailout, but 1) they couldn't have refused it 2) they paid it all back and 3) that's not a reason to regulate Citi more, rather, to not bailout big banks.  The solution to bad regulation isn't more regulation, but less. It's not only creates more efficient incentives and allocation of risk and capital, but is far easier to implemented.

For example, the latest regulatory proposal actually targets spreadsheets:

The regulator stated that insurance firms will be expected to demonstrate appropriate controls with regard to key internal data flow systems such as spreadsheets. These controls should take the form of, among others, input validations, change and release management, disaster recovery and documentation.

They might as well validate my Post-it notes too.

Back to Citi, I think the best way to increase value there is to break it up into into manageable pieces. One simply can't manage something that big, as the big banks are demonstrating. 

Bad Regulations Don't Go Away

A poor guy in Iowa was dismissed via an absurd no-tolerance policy by the FDIC:

A 68-year-old Des Moines man fired from a Wells Fargo call center for putting a cardboard dime in a washing machine in 1963 has been cleared to return to work in the banking industry, the Federal Deposit Insurance Corp. said. Wells Fargo is under no obligation to rehire Richard Eggers, however... Wells Fargo said it was simply complying with the regulations, which carry a $1 million-a-day fine... A 1950 federal law prevents FDIC-insured banks from employing workers who have been convicted of a crime of dishonesty or breach of trust. In 2008, Congress passed a law that forced mortgage loan originators to perform similar employee background checks.

This dumb law was overturned, but it has to be this dumb, take 62 years, and merely generates an exception not a rewrite.

No-tolerance policies are primarily adopted by institutions when people have no faith in discretion. As Phillip Howard has argued, discretion dominates rules because reality is always more complicated than any rule contemplates. Yet by law bad outcomes can be legal under discretionary frameworks, so lawmakers like zero-tolerance policies because in theory that eliminates the perceived problem.

It's a classic example of the perfect being the enemy of the good. Anyone really enthusiastic about regulations fixing rather than causing problems in the financial industry can't rely on history, as financial regulations have been and continue to be generally irrelevant. We have one set of regulators who look over hard copies of all our traders personal stock trades, presumably to look for front running.  This is silly for two reasons. First, we don't have retail flow like a brokerage, so we would be hurting our company for personal gain in this case, and clearly we have a greater incentive to stop such individual malfeasance than regulators. Secondly, looking at literally hundreds of thousands of company trades, and comparing them to a stack of personal trades in hard copy (with various formatting), is like looking for a risk premium using covariances, a waste of time.

Vol of Vol

I was at this year's OptionMetrics conference and it was a nice overview of 13 different papers. One that stood out was Guido Boltussen's paper (coauthored by Van Bekkum and Van Der Grient of Erasmus U) on Unknown Unknowns: Vol-of-Vol and the Cross Section of Returns. The idea was simple. Instead of sorting by vol, they sorted by vol of vol, and generate a rather large annualized return  difference (10%) between the high vs. low vol-of-vol buckets. It's a bit like the guys at Gillette coming out with 5 blades, applied to volatility. Here's the standard graph with monthly excess returns:

In any case, it's intriguing. I haven't looked too hard at this, but it's a pretty obvious extension of the well-known low vol premium, and plan to investigate further. Clearly vol-of-vol is correlated with vol, so it would be interesting to see how these variables relate to each other in explaining the now well-known low-volatility premium. Perhaps volatility is a poor man's estimate of the 'true' factor, vol-of-vol, or vol-of-vol-of-vol.

There were several interesting papers, and I noticed that as usual, the risk premium (aka 'price of risk) was usually negative. Now, the risk premium should be positive, a premium, but empirically is usually negative; I don't think that word doesn't means what they think it means. I think it's better to say, here are excess returns, controlling for various obvious alternative characteristics we know to be correlated with excess returns. 

OptionMetric User Conference Monday

The 2012 OptionMetrics Users Conference is Monday, October 15, at the New York Society of Securities Analysts Conference Center, Times Square, New York. I'll be speaking Monday at noon, but there are 4 different sessions each with 3 speakers discussing various strategies and tactics.

 See details here. 

Ego and the Low Vol Premium

In my book The Missing Risk Premium, I have a chapter on why if anything there appears a negative risk premium: more risk, lower return. It's obvious in lotteries and gambling, where the most improbable events have the lowest expected returns, but true in less trivial areas such as options and stocks. One of the causes is signaling, that investors tend to want to show others how awesome they are by trading frequently just like a real alpha generator would.

 It's interesting to note that in games there's a profound dichotomy between the optimal tactics for beginners and experts. For example, Simon Ramo notes that among the very best tennis players, to win you need good winning shots; to be a good average player, you need to merely lower your failure rate. In expert tennis, 80% of the points are won, while in amateur tennis, 80% are lost. The same is true for wrestling, chess, and investing: beginners should focus on avoiding mistakes, experts on making great moves.

Yet if the distinguishing characteristic of an expert investor is whether they are being aggressive, then any aspiring expert is forced to be aggressive because this signals to others that he truly is an expert, and finance is all about getting other people to give you money to manage. Thus it should come as no surprise that if you give people advise to invest in simple index funds or to focus on low volatility stocks because you can do little damage, and save a couple percent a year, many will dismiss this advice. The favorites of aspiring financial moguls are volatile, because one isn't going to hit a 'ten-bagger' on Coca-Cola (vol of 15%), but rather Netflix (vol of 70%).

 An example of how investors think about the perceptions of others was nicely demonstrated by Baumeister, Heatherton, and Tice in their 1993 article When ego threats lead to self-regulation failure: negative consequences of high self esteem. They gave subjects a video game involving flying a plane through obstacles, and after 20 minutes of practice (and secret recording), they were told they would receive $2 if they matched a time set to just below their actual average time, but $4 if they did some bit better than their average time. Subjects didn't realize this criterion for payout was scaled to their individual prior performance, but rather, thought they were from population averages (ie, those dopey other guys).

Half the subjects were randomly assigned to the ego-threat condition. For these subjects, the experimenter added the following remark: "Now, if you are worried that you might choke under pressure or if you don't think you have what it takes to beat the target, then you might want to play it safe and just go for the two dollars. But it's up to you."

 When that little statement added, more then took the gambit and did worse on average.

Our desire to impress others causes us to take too much risk. On the bright side, this implies some rather simple strategies like low volatility investing, because I don't see it going away.

Guilt Dominates Shame

I stumbled across a fascinating lecture on shame and guilt by June Tangney, a psychology professor at George Mason. I used to think they were the same, but she points out they are profoundly different.

She starts with a story about when she saw her little daughter kick her sibling, she wanted her daughter to feel bad. That is, though she loves her daughter, she wants her snowflake to feel bad sometimes, as it's a healthy corrective when one inevitably does bad things. The question is, how should you feel bad about yourself in such cases?

Guilt and shame are emotions not present at birth because they require a sense of self and standards; babies don't feel them. She defines shame as feeling bad about oneself, guilt feeling bad about behavior. Shame is feeling 'I am bad', guilt is feeling 'I did a bad thing.'  Both are painful, but guilt is not nearly as overwhelming.

 Guilt leads to thinking about behavior and its effect on others, shame is more focused on how others think about us. Shame attempts to hide and deny to escape the situation, even anger.  Guilt involves regret and remorse, and leads to a focus on reparation and redress. Guilt and other-oriented empathy go hand in hand, while shame interferes with empathy.

As a good scientist, she looks at data and finds people more prone to shame are not as mentally healthy as people prone to guilt. Further, people prone to shame blamed others more for their misfortune, classic defensive Freudian projection. Shame-proneness is more related to depression, anxiety disorders, low self-esteem, drug use, destructive behaviors, etc. Guilt proneness, in contrast, is associated with lower recidivism by convicts.  It doesn't seem to be inherited, as the correlation between parent and child in shame-proneness was a measly 0.1.

She recommends guilt as the moral emotion of choice. Parents should minimize shame and humiliation, but rather to call children's attention to the harm that they've done, make redresses, and empathize with those they have hurt.  Guilt is much more constructive and proactive, and changing behavior is the key to improving one's situation.

Confirmation of Low Vol Dominance

The Institutional Investor takes on the Low Vol effect. Their simulation for US stocks, where they used the top 3000 tickers, and created a LoVol portfolio using the bottom 20% ranked by volatility, and HiVol the top 20%. Here's the results:

January 1974 through November 2011, LoVol (portfolio) beats the index by 59 basis points a year on a compound basis... As for HiVol, its performance relative to the index is truly dismal: It underperforms by 592 basis points a year on a compound basis.

This helpful nugget is nested in a theory that defies comprehension: reconstitution. My brain hurts trying to conceive how this works, because they say things like

If we hold a stock that drops out of the universe, we will certainly be selling it at a relative loss, because the stock has failed to keep up with the other stocks. After all, it has lost its place in the top 1,000. In essence, the strategy requires that we sell low in this case, creating a drag on performance.

But in this example, the stock had already lost value; selling it in backtests creates no further loss. They then note:

When multiple stocks are combined in a long-only portfolio, their interaction actually causes the portfolio to have a higher compound growth rate than the weighted average compound growth rate of the stocks in that portfolio.

The authors note this "can be quantified using highly technical mathematics", and for a proof that E(x+y)>E(x)+E(y), it must be highly technical indeed.  But why bother with a 100 page proof using the Taniyama-Shimura conjecture when you can simply go long SPY, short its constituents, and laugh all the way to the bank. Arbitrage! Methinks they are confused.

 It's good to know the why, because without a good theory for what is driving the low vol effect, one is reasonably skeptical, and of course I have one in The Missing Risk Premium: Why Low Vol Investing Works. Remember: theory without data is bullshit, and facts without theory are trivia.

Poor Ina Drew

Executives at large financial corporations tend to be smarmy bureaucrats, not savvy investors. They rise to the top via excellent credentials, good connections, better timing, and learning to not say anything so clearly as to betray their ignorance on things large and small. Ina Drew sounds like a Dilbert caricature.

A New York Times puff piece on the former JPMorgan CIO tries to paint her as the scapegoat of their infamous first quarter debacle, but their description of the $6B bad trade/hedge sounds like she never understood it. Most of the piece is about her life, but then they get to The Trade and there's vague and inconsistent descriptions of events given the author's access. They mention

Drew’s deals essentially turned on one key question she seemed to answer correctly more often than most (or at least when it mattered most): Would interest rates go up or down?

I find this laughable. Anyone who's been in finance for a couple decades knows that anyone with an edge in interest rate forecasting is deluded, ignorant, or not a corporate executive. That's just not where those kind of people end up. An executive who thinks calling market direction is really important doesn't understand the Serenity Prayer, that banking is about intermediation and not speculating.

 As a short credit position got large in 2011:

Ina knew the product, the size they were trading, but she did not know what the true P.& L.” — profit and loss — “impact could possibly be in a stressful scenario,” he said...One serious defect in the risk evaluation of Iksil’s position was that its limit was folded into the aggregate risk of the unit’s entire portfolio. In other words, Iksil could continue to increase the position without triggering alarms.

How can a CIO not understand what would happen in the mild hiccup in the first quarter of this year? How could such a large position be 'folded into the aggregate risk of the unit' yet be marked to market?  I suspect she understood it as so many senior executives do, which is, not very well. They then discuss how they tried to close the position by putting on a correlated position going the other way. Clearly this didn't work as expected, which means it was a bad hedge, clearly in the CIO's bailiwick. 

They mention she's became quite wealthy. She should be grateful to the Gods of corporate chance.

Short Sales as a Coasian Solution

The Coase theorem states that if there are no transaction costs, bargaining will lead to the same outcome regardless of the allocation of property rights. This highly counterintuitive result has a very profound implication, such as that the law should focus more on predictability than justice, because if one suffices with a reasonably just law that is highly predictable, one should expect the same net effect in terms of people and businesses doing what they would have done to maximize their welfare if the law were perfectly just. Predictability is a practical supreme objective for any justice system.

The many steps a bank must take in a foreclosure, and various rights of homeowners to supersede a foreclosure sale, basically make the abandoned property the concern of no one for a long time (avg time of delinquency: 631 days!), and these laws are counterproductive responses. For example, during the crisis a new law allows mortgage holders to reclaim the house within a 6 month 'redemption period', which is designed for the possible but improbable case where someone falls behind on a mortgage but has positive equity in their home. In practice just means the property lies fallow for 6 more months than necessary.

 Short sales have become more popular because they can shorten the time a house is not being used. In a short sale, the mortgage holder sells the home for less than their mortgage outstanding (banks lose), but the property is then owned by someone who actually will pay taxes and mortgage payments and take care of the property. The benefit to the borrower is that this is not as harmful as a bankruptcy or foreclosure on their credit history; the benefit to the bank is the property avoids wasted time with no upkeep, and the house can actually start generating cash-flow again.

It's a second best solution forced on us via naive regulations, but anything that shortens the time properties remain unkempt, unpaid, the better.

Straw Man Argument

From a pro-Obama video quoting Obama speeches:

 Obama: "You would think (my critics) would say maybe some rules and regulations are necessary to protect the economy..."

 The idea that the debate is between no regulation vs. good ones is a caricature, but it's the conviction of many. These people have no understanding of how many different rules exist, and that there's a difference between less regulation and zero.  As regulations proliferate it's impossible to enforce them all fairly, and so enforcement becomes selective and capricious. Legislative micromanagement is the road to crony capitalism, why so many senior risk managers at large private institutions are ex- government officials with some connections with the regulators. 

Fixing Health Care

Listening to the presidential debate on how to insure everyone's health care without penalize those with higher expected medical expenses is amusing because it's redistribution masquerading as some simple engineering problem. It's always more efficient in these scenarios to simply redistribute, rather than play games so no one sees this as what is really going on, but then this is a harder sell and thus the equilibrium. I was struck by this line in David Henderson's review of Priceless by John Goodman:

Strong evidence for Goodman’s view that there are good effects from having consumers face real prices for health care comes from the area of cosmetic surgery. Such surgery, he notes, is rarely covered by insurance. He points out that, uunlike in most areas covered by insurance, patients can typically find a package price that includes all services and facilities and compare prices prior to surgery. Moreover, he notes, prices adjusted for inflation have fallen over time as technology has improved. He notes that for the kinds of surgery covered by insurance, improvements in technology are blamed for rising prices.

I'm reminded of the saying that things are never so bad that they can't get worse, which is where I see us going. Less regulation seems improbable regardless of who wins the election.

A Deep Understanding

 

There's an interesting article on Slate by someone miffed at all those 'correlation doesn't impy causation' slams common in retorts.  He makes the good point that, while not proof, correlation is suggestive, and consistent with causation.  If you want to be a pedant nothing non-tautological can be proven (see Hume's problem of induction). It reminds me of the saying 'absence of evidence is not evidence of absence.' This is wrong. It's not proof of absence, but from a bayesian perspective, it should increase one's belief in absence.

More Evidence that Envy Trumps Greed

Two Israeli economists, Haim Levy and Guy Kaplanski, have a new paper, Investment Choices with Envy and Altruism. The go over what kind of utility functions can rationalize the following preferences.
They asked people about the following scenarios:

 1.The value of the subject’s investment in stocks increases during the year from $25,000 to $27,500, and the stock index also increased at the same rate.
2. The subject’s investment and the stock index increase exactly as in Scenario 1, while the portfolio of the peer group appreciates by 50%, from $25,000 to $37,500.
3. The subject portfolio as well as the stock index decrease during the year from $25,000 to $22,500.
4. The subject investment and the stock index decrease exactly as in Scenario 3, while the portfolio of the peer group depreciates by 50%, from $25,000 to $12,500.

 Needless to say, they find that "envy dominates the results as the utility generally decreases when the peer group earns more and it increases when the peer group loses more than the subject." It's relative wealth that matters more than absolute wealth.  This is consistent with the theoretical thesis of my book, The Missing Risk Premium (now with 5 reviews, and the 'Look Inside' feature is now enabled!).

Bank Stock Timing Strategy

My bank vega theory I alluded to yesterday argued that when banks are weak, they are in a negative vega zone: the potential for getting stopped out makes them want to avoid risks, not take them. For fun, I looked at a trading implication, based on the idea that I alone, via my theory, know when banks will be timid or growing.  A timid bank doesn't make money, so it's best to avoid it then. It's not short term enough for it to be really useful, so I'm sharing, and I think all investors thinking about their bank index allocation should find this interesting.

A calculated a backward-looking correlation of the past 252 days of daily bank returns with the changes in the VIX index, because Ken French's website has daily returns for a bank index going back to 1986, I have VIX data from 1986, and I can update this with the KBX bank index for more recent daily data.

I then take this correlation, and find its median. I then say, invest in the SP500 when this correlation is high in absolute terms (very negative, all correlations with the changes in the VIX are negative), I don't want to invest, because the banks won't be investing and making money. In the 50% of the time that correlation is low (again, in absolute value), I'm in the banks. Here's a comparison of that strategy, and clearly timing based on the bank-VIX change correlation outperforms simply being always long banks.

Here's what it looks like in Scatter-Plot, comparing the future 3-month return in the two regimes.  The black vertical line in the center is my best stab at putting a median in there.

We are currently at -70% for a correlation, still bad for banks (median around -62%). This is all out of sample, and a pretty simple rule. I did try this with Ken French's other industries, and found no similar pattern.