4 posts tagged “finance”
The intelligent and thoughtful Felix Salmon makes a subtle and interesting error--an error that I would make on at least a monthly basis had Robert Waldmann not patiently explained all this to me in the winter of 1986--in discussing Kelly risk analysis. Pushing leverage beyond the Kelly point does not decrease expected return. Rather, it decreases the likelihood of organizational survival and the chance that you will be wealthy. If you are acting as one of many agents for a well-diversified principal, you will in general want to ignore the Kelly point and leverage yourself up to the gills. If your objective is, instead, to maximize your own chances of remaining in the game with boasting rights, you will position yourself at the Kelly point.
A stark way of seeing this difference is to think of the following situation: Matt Rabin from the office beneath mine comes up the stairs and offers me the following: I start with a stake $1. I can wager none, some, or all of my stake. He flips a fair coin. If it is tails, I lose my wager. If it is heads, I win twice my wager. We do this ten times in a row, with my stake growing or shrinking.
To maximize expected return--this is, after all, a very advantageous game for me--I should be my whole stake, and let it ride time after time. After 10 rounds, there is one chance in 1024 that I have $59,049 and 1023 chances in 1024 that I have zero, for an expected portfolio value of $57.67.
The Kelly point, by contrast, says that I should wager 1/4 of my current stake each round. If Matt flips ten heads, then I have only $57.67 instead of $59,049. And my expected final wealth is only $3.25 instead of $57.67. But my median final wealth is not $0 but is instead $1.80. I make money not 1/1024 of the time but 638/1024 of the time. And if Robert were here he could prove in five lines that as the number of rounds goes to infinity an agent wagering according to the Kelly criterion almost surely ends up wealthier than an agent choosing his wager from his or her stake according to any other rule. The Kelly point makes sense if you are risk averse (and if this portfolio is a major component of your wealth) or if organizational survival and relative organizational prosperity is your major goal.
If Matt showed up at my office and announced that we were going to play this game on each of the next 1024 days, I would have no trouble choosing to follow the bet-the-limit strategy rather than the Kelly strategy on each day. 1024 x $3.25 is only $3328, which is a lot less than 1024 x $57.67 = $59049. But what if Matt says that this is my one and only one day? The right way to think about it is that my marginal utility of wealth is surely pretty flat over the range of $50,000 or so, and so I ought to be risk-neutral in this particular situation. The right way to think about it is that I "buy" lots of lottery tickets of various types during my life, and that the right strategy is to maximize the expected value of each lottery ticket--not to apply the Kelly criterion to each situation individually. (Of course, the generalized Kelly criterion--maximizing the expected value of the log of your portfolio--for 1024 rounds is not to apply the Kelly criterion to each round independently.)
But I would find it hard. I would have a hard time giving the 1/1024 chance of winning $59049 its proper weight in the face of the 1023/1024 chance of suffering the humiliation of bankruptcy.
In the end, however, I would be the limit. The humiliation for an economist like me of violating the axioms of expected utility is much worse than the humiliation of losing my entire stake.
Kelly Criterion finger exercises at: http://spreadsheets.google.com/pub?key=p_zylRhg4towI71xZsP62Fg
At night in the suburbs of San Francisco, some of us awake as the hills echo and re-echo with the howls of the coyotes that have fed well on Glenn Rudebusch's chickens. We then lie awake, worrying. We worry why the Great Moderation in the U.S. business cycle on the real side that we have seen since the mid-1980s has not carried a big reduction in financial-side variability with it. We toss and turn, worrying that the real-side volatility decline has been part good transitory luck and part statistical illusion, all because people in financial markets putting their money where their mouths were do not project the continuation of the Great Moderation into the future.
Christina Wang's paper lets us sleep more easily, even if the coyotes continue to prey upon the chickens of Federal Reserve Bank Vice Presidents. It teaches us an important and valuable lesson: a financial system that is doing a better job will be highly likely to have both higher financial and lower real volatility.
When a firm goes bankrupt and defaults on its debt, it may be because it has had bad luck, it may be because it was badly managed, or it may be because it suffered from moral hazard--took account of the fact that in the lower tail the losses are eaten not by the firm but by the bank that loaned it the money. Banks that have a hard time distinguishing between these possibilities will be averse to lending--charge a high interest rate premium on loans--to firms seen as having a high degree of undifferentiated idiosyncratic risk. Improvements in data collection and analysis that allow firms to differentiate will cause banks to fear undifferentiated firm-level idiosyncratic risk less, and charge lower interest rate premiums for such lending. Other things being equal, firms will smooth production more, and smooth cash-borrowing requirements less, seeking to squeeze out more productive efficiencies by taking on more financial risk. To the extent that improvements in data collection and analysis reduce banks' fixed costs of monitoring loans, other things being equal banks will do more to diversify away firm-level idiosyncratic risk.
When a bank goes bankrupt and defaults on its debt, it may be because it has had bad luck, it may be because it was badly managed, or it may be because it suffered from moral hazard--took account of the fact that in the lower tail the losses are eaten not by the banks' shareholders but by those who hold or guarantee its liabilities. Improvements in data collection and analysis by those to whom banks owe their liabilities will allow them to better classify banks, and so the cost to banks of portfolios with bank-level idiosyncratic risk will fall. Other things being equal, banks will be willing to take on more bank-level idiosyncratic risk.
Of course this function that Christina Wang identifies is the primary job--one of the primary jobs--of financial markets: to diversify away idiosyncratic risk, as was ably explicated by that notable predecessor of Lintner and Markowitz, William Shakespeare. As Shakespeare writes, Antonio, the Merchant of Venice, does not fear that the lower tail of his portfolio return distribution extends far enough down to the state in which his heart is cut out with a knife. Antonio he has a properly-diversified portfolio. The banker lending him the money uses the highest information technology of that day: wandering down to Venice's Grand Canal, loitering on the High Bridge, and gossiping. The banker concludes that Antonio has:
an argosy bound to Tripolis, another to the Indies; I understand moreover, upon the Rialto, he hath a third at Mexico, a fourth for England, and other ventures...
Here the analogy breaks down. Negative transitory systematic news does indeed provoke a crisis in Antonio's affairs, but he is rescued not by a competent, technocratic lender of last resort but by his bride disguised as a teenage judge.
Christina Wang hopes that starting sometime in the mid-1980s we took a jump toward the ideal financial world in which one of CAPM's cousins holds, in which idiosyncratic risk is not priced because it is properly diversified away, and in which as a result the real economy can grab for all the production-smoothing efficiency benefits without worrying about firm- or bank-level costs of default or illiquidity. This shift could drive a reduction in real-side volatility coupled with an increase or no change in financial-side volatility.
She has a nice theoretical costly-state-verification model of the effects of improved data collection and analysis technologies. She has a very interesting theoretical Dixit-Stiglitz-based three-period model of the joint determination of real and financial volatility. The key insight is a very good one: that production-smoothing has not just manufacturing-side and labor-side efficiency benefits but financial-side efficiency costs: only if banks are confident in their ability to monitor firms and large depositors confident in their ability to monitor banks will firms be able to easily and cheaply borrow the money they need in recession to enable a production-smoothing corporate strategy. The fact that times of recession are times when a firm's free cash is likely to be uniquely valuable and not to be best invested in building up inventories is a potentially powerful explanation of why we have, historically, seen the reverse of production-smoothing in the American economy. She has interesting empirical results that suggest that banks and firms have reacted to a likely information-driven fall in the cost of idiosyncratic financial risk to take on more of it. The theory is sound and convincing. The micro empirics are interesting and suggestive.
But how much can this channel add up to on the macro level? How, exactly, does ICT help bankers? Working for the original J.P. Morgan, Charlie Coster was on the boards of 88 railroads at the turn of the last century and died of overwork--Morgan is reputed to have recruited Coster's successor while they were together carrying Coster's coffin to its grave. What would today's ICT have done to increase Coster's contribution to Morgan's bottom line, exactly?
And how much of the Great Moderation in real-side economic volatility can this channel account for? Recall the size of the Great Moderation: a 40% fall in the standard deviation of the cyclical component of GDP, more or less the same however you choose to measure it. A fall in spite of the fact that technology and cost shocks have in all likelihood been quantitatively greater in the past ten years than in any other post-WWII decade save possibly the 1970s.
As Christina Wang says, her paper as written can't do the job. It can only do about a third of the job--although Doug Elmendorf said half last hour. The model as extended quite possibly could.
In this literature, the game that is being hunted is the positive correlation between production and inventory investment that we saw in the past. In a standard production-smoothing model inventory investment should be relatively high when production is relatively low, and sales are very low. Instead--back before 1985--inventory investment was high when production was high. This shift could be possibly traced to Christina Wang's mechanisms. But it can account, in my back-of-the-envelope guess, for not a 40% but a 15% decline in the standard deviation of the cyclical component, whatever that is.
The big game for this model--as Chistina Wang says in her conclusion--will, I think, come from applications of models like this to the household sector. It's not just firms that have benefitted from the application of information technology to credit screening. I have gotten three offers of VISA cards and two offers of what were described as "guaranteed low interest" home-equity loans so far this week. Plus the people behind the counter at my most local Starbucks have started asking me if I'm interested in a no-annual-fee Starbucks VISA that will come with $25 of free caffeinated drinks. I don't know whether they are doing this to everybody or whether there is something special in my file. The smoothing-out of household durables purchases will, I think, be an important part of the Great Moderation when we finally nail it down. And I think that's where the high returns from Christina Wang's model will come.
Last, the smoothing out of residential construction--if it indeed stays smoothed-out--may well turn out to be the heart of the matter. One branch of the conventional wisdom is that the smoothing-out of residential construction is a result of good luck that is about to end: that America's banks have been offered too much rice wine by the People's Bank of China, and have responded by lending like drunken bankers: $600,000 zero-down floating-rate loans to single-earner middle-class families buying three-bedroom houses in Vallejo, CA: and we will be sorry.
Christina Wang's paper suggests a second possible explanation. That recent residential investment financed by so-called "non standard" mortgage loans is a result at least in part not of the inebriation of the banking sector but of the ability to more finely calculate risk and return than was possible in the days when your mortgage had to be 30-year-fixed, 20% down, with amortization plus real estate taxes amounting to no more than 33% of last year's household income. That was an inadequate screen. What, really, are the current screens? How good are they? The application of models like this to residential financing may be the real big game here.
Megan McArdle thinks that Ken Lay and Jeff Skilling and Richard Causey (who in the end pled guilty just before the trial, thus playing havoc with Skilling's and Lay's planned defenses) were factually innocent. Or maybe she doesn't:
Asymmetrical Information: the only solid evidence that Messrs Skilling and Lay were "just as guilty" as Mr Fastow comes from... Mr Fastow, who got a drastically reduced sentence for telling the prosecutors this. Mr Fastow is the fellow we know stole from the company; the conviction of the others is based largely on his testimony. I haven't followed the case all that closely....
There is a small (but I believe growing) school of thought on Enron which argues that Enron was a fairly healthy business--done in not by deteriorating financials, papered over by accounting fraud, but rather by a collapse in trust in the market place, helped along by appallingly bad financial journalism, which basically caused a bank run. Since Enron was a trading operation, which requires liquidity to stay in business, profitability was irrelevant; they were killed by a credit crunch.
I don't assert that this is the case, mind you, though I find it at least plausible given my knowlege of the case.... Mr Fastow is the only higher-level executive against whom the government possessed a smoking gun. So this isn't a case of Fastow getting the ride on the lifeboat because he was the first conspirator to break, which is the general excuse for such plea bargains. In this case, the most obviously, incontravertibly guilty person in the case got a much, much lighter sentence because he testified against others whose guilt was more arguable....
According to Enron, it had stockholders' equity of $11 billion at the end of 2000. It gained an extra $2 billion from the California energy crisis in the winter of 2001. Yet by the fall of 2001 all of that was gone--Enron was bankrupt.
A bank that runs into confidence problems--that has a "run on the bank"--can go bankrupt. A bank has lots of short-term liquid liabilities, and lots of long-term illiquid assets that it can turn into cash only at heavy discounts.
But Enron was not a bank. Enron was--or was supposed to be--an asset-light trading operation. Its assets and liabilities were (a) securities and (b) promises to buy or sell energy and other commodities in quantity at various times in the future. These are not the kinds of assets that should lose much value even in a fire sale--even if it is accompanied by "appallingly bad financial journalism."
When Enron lost credibility--if it were a trading company--it should have been able to unwind its operations coherently, sacrificing some but not all of its $13 billion in equity to grease the deals. It was--or was supposed to be--a properly-hedged trading company with liquid assets, not a highly-leveraged bank with illiquid assets. But Enron went belly-up. Its unsecured creditors got 30 cents on the dollar. They are now (with Andy Fastow's help) hunting for the heads of the banks that did the paperwork for Enron's financial frauds.
It seems overwhelmingly likely that the $13 billion in stockholders' equity never existed. Enron marked its positions to market--which means that when it signed a contract with, say, a midwestern city to provide it with power for the next 20 years, it claimed the whole value of that contract as income in that quarter. And how did it calculate the long-run present value of the contract? It let the person who negotiated the contract say what it was worth.
Enron had--before Skilling--had a somewhat more sophisticated risk-control system. They had people in charge of reviewing and monitoring the contracts made--experts in figuring out what the contract would be worth if it were a liquid security. People like Vince Kaminski, who one day, Kurt Eichenwald reports, got a call from Enron President Jeff Skilling:
"There have been some complaints, Vince, that you're not helping people do transactions," Skilling said. "Instead, you are spending all your time acting like cops." A pause. "We don't need cops, Vince."
When you remove the "cops," you create the mother of all principal-agent problems. Each time a contract was completed, the contract was valued as if it were a liquid security by the negotiator who closed the deal. This meant that a negotiator desperate to reach a deal could (a) give away the store to the counterparty, (b) still make assumptions in his "mark to market" valuation that made the deal look really profitable, and so (c) help Skilling and Lay report another banner quarter for earnings. But you are reporting profits that do not really exist--and claiming stockholders' equity that is not really there.
For example:
http://www.thunderbird.edu/wwwfiles/pdf/about_thunderbird/case_series/a15040016.pdf#search=%22arthur%20anderson%20enron%20restatement%20of%20assets%22: The Dabhol, India, project was a microcosm of Enron’s business model. In December 1993, after roughly 18 months of negotiations, Enron Development signed a long-term power sales contract with the electricity board of Maharashtra, India (MSEB). Enron would build a 2000-megawatt natural gas- fueled power plant at an estimated cost of $2.8 billion. The Maharashtra energy board agreed to buy 90% of the power produced by the plant at a U.S. dollar-denominated price for a minimum of 20 years.
From the very beginning, problems mounted. It took Mark more than two years and millions of dollars in negotiations, court cases, and contract restructurings to get the financing in place. Within months, a change in the Maharashtra government resulted in an outpouring of project opposition. Once again, Mark went into overdrive for months of negotiations and renegotiations. Finally, on February 23, 1996, a new contract was signed between Enron and the MSEB. By December, financing was in place and the project’s construction resumed. But opposition continued, and it became clear that the MSEB would never be willing or able to pay for the power (estimated at over $30 billion for the project life).
The project was dead. In the words of one Wall Street Analyst on Enron’s India activities, “I’ve never been to another country where every single person hates one company.” Phase I of the project was operational for only a short period of time, and Phase II, as of December 2003, was still only 80% complete. The Dabhol power plant today is considered a failure for everyone involved but Rebecca Mark. Mark earned bonuses for both the original deal [in 1993] and the successful renegotiation [in 1996]. In 1996, Enron International generated 15% of Enron’s total earnings, and was expected to grow at double-digit rates for years to come. Rebecca Mark was named CEO of Enron International. Many people believed that Rebecca Mark would be Enron’s next CEO...
And:
[I]n July 2000 Enron signed a 20-year agreement with Blockbuster Video to introduce entertainment on demand.... Pilot projects in Portland, Seattle, and Salt Lake City were created to stream movies to a few dozen apartments.... Based on these pilot projects, Enron... recognized... profits of more than $110 million from the Blockbuster deal...
And:
Enron entered into a $1.3 billion, 15-year contract to supply electricity to the Indianapolis company Eli Lilly... [booked] the present value of the contract... more than half a billion dollars, as [current] revenues. Enron then had to report the present value of the costs of servicing the contract as an expense. However, Indiana had not yet deregulated electricity, requiring Enron to predict when Indiana would deregulate and how much impact this would have on the costs of servicing the contract...
And we haven't even gotten to Andy Fastow and his special purpose entities yet...
Stewart Hamilton and Alicia Micklethwait (2006), Greed and Corporate Failure: Lessons from Recent Disasters (London: Palgrave: 1403986363), has the best short thumbnail discussion of Enron (and of other things) I have yet found:
"At the end of the year [2000], proposed write-offs for asset impairment were on the order of $7 billion against reported shareholders' equity of $11.4 billion. Such a write-down would have greatly increased Enron's debt/equity ratio... loss of investment-grade status... the trading business would be destroyed. Somehow, Fastow persuaded Anderson to defer consideration of this until after the year end. Furthermore, the broadband venture was losing money... the fall in the value of Enron's share price was likely to trigger its guarantee obligations in relation to many of the SPEs..."
