It’s unfortunate that people don’t like to talk about trades gone sour.
Behavioral economists and market observers have known for years that people take credit for their success, but the same people attribute failures to other people’s behavior or advice. In addition to people’s reluctance to face their own mistakes, much less talk about them, the media is not much help. Efforts to find scandals are all too easy; they’re the lazy man’s reporting. The result is that one can expect more stories of victims than analysis of mistakes. That’s unfortunate because other people’s mistakes are a possible source of insight; besides they’re usually cheaper than one’s own.
The Goldman Sachs story references in PART 1 may prove to be an exceptional source of information on mistakes to avoid. But, even when the facts do surface, reporters’ search for villains may obscure the potential lessons. GS clearly doesn’t come out looking like a saint. But, at the same time, it seems silly to portray them as a beneficiary. As mentioned in PART 1, loans from Buffet and TARP as well as big write-downs of assets aren’t typically what one would associate with a winning portfolio. So, how did GS and others end up needing the help?
As background, the reader will find a lengthy quote below. The quote isn’t included to explain how organizations wrecked themselves. It talks about millions which is small change when considering the scope of what happened. Rather, the quote will serve as a source of illustrations.
Also, the quote was taken from a story making a bullish case for GS stocks. However, PART 1 already made the case for avoiding investments in situations involving conspiracy stories. Thus, no reader who checks out the full story could misinterpret my reason for the quote. The full story can be found at:
http://online.barrons.com/article/SB127146775162378679.html
“The SEC complaint shows the lengths to which Paulson -- and Goldman -- went to create a CDO that Paulson figured stood a good chance of collapse.
Paulson & Co., led by John Paulson, correctly anticipated that the worst subprime securities would come from adjustable-rate loans to borrowers with low credit scores in such states as Arizona, California, Florida and Nevada, where home prices had soared, the SEC said.
Paulson & Co. cleverly sought to bet via a CDO, or a collection of existing sub-prime securities, rather than simply bet on a newly created pool of loans. The securities that were the focus of Paulson's efforts were rated Triple-B by the rating agencies and stood beneath a large group of highly rated bonds.
The Abacus CDO was backed by about 90 individual Triple-B tranches from sub-prime deals. Paulson probably recognized that if losses on the underlying subprime pools hit 10% to 15%, the triple-B tranches would be wiped out, resulting in a total loss on the Abacus CDO despite Triple-A ratings on the instrument. That is indeed what happened.
The losers from the Abacus deal were a German bank, IKB (IKB.Germany), which lost $150 million, and Royal Bank of Scotland (RBS), which lost $840 million.
The Scottish bank's loss, which helped lead to a giant bailout by the U.K. government, itself is fascinating. ACA Capital, designated to select the underlying securities, originally guaranteed the $909 million top-rated tranche of the Abacus deal for just 0.50%, or about $4.5 million.
ACA subsequently got buried by mortgage losses and couldn't pay the claim. ABN Amro, the Dutch bank subsequently bought by Royal Bank of Scotland, provided a backstop to ACA on the deal for an estimated $2 million. That $2 million guarantee ended up costing Royal Bank $841 million when the CDO collapsed, and most of that money was paid to Paulson.
Royal Bank's huge loss raises questions about whether the ABN Amro managers knew what they were getting into and demonstrates the risks in the financial-guarantee business, which Buffett has described as ‘picking up nickels in front of a steamroller.’
In a statement late Friday, Goldman emphasized that it lost $90 million in the transaction, that IKB was a highly sophisticated investor and that ACA had ‘every incentive’ to select appropriate securities because it issued a $900 million guarantee on the deal.”
The quote is rich in examples of mistakes. There are mistakes both involving actions taken and interpretations of the actions. As written, it tells an interesting story. But, when one looks at it from the perspective of an analyst rather than a narrator, it’s always good to start with the most important item.
1) Asymmetric payout as a warning
The most important point is to be extra careful when asymmetric payouts are involved. From the perspective of mistakes to avoid, Buffet’s quote comparing the risks in the financial-guarantee business to "picking up nickels in front of a steamroller” is the most important point.
Why is this point most important? After all, if payouts on good trades verses losses on bad trades are markedly different, of course one would be careful. One would do the math on the payouts, calculate the odds they imply, then compare them to one’s own best guess as to what the odds really are. What’s the big deal?
First, one should remember that if the payouts are very different, any small error in the estimate of the odds will have a big, I mean BIG, impact on the payout. That in itself is a big deal. But it’s worse than that, never mind estimates. If there are even small errors in how the odds are measured , and there are aways some errors, disaster can follow.
Second, there is considerable evidence that people do funny things when large sums of money are involved. It can be seen as irrational or as evidence that a million dollars means something other than one dollar times a million. Economists say the marginal utility of money isn’t linear. Doesn’t mater what one calls it; lottery ticket sales clearly aren’t in the financial interest of the buyer from the perspective of the probable payout verse the price. Clearly, there is more to the phenomonal success of lotteries than just people shelling out a buck for a few days of the dream. It doesn’t end there. It also seems that most people don’t do extremely large numbers well. Witness how many don’t distinguish between million, billion, and trillion when discussing public policy. So, it seems important to be especially clear about what one is buying and to not confuse dollars and dimes.
Third, often asymmetric returns are associated with what is known as tail-risk in risk management and financial economics. Tail-risk involves very low probability events. This is exactly the area where traditional quantitative financial economics tends to fail. Further, the reason it fails seems to reflect basic, or at least common, perceptual predispositions of humans. We are so dependent on basing our expectations on what is common that we underestimate the probability of the uncommon. In quantitative financial economics, this surfaces in assumptions about the probability distributions associated with events. So, one has to overcome one’s own predisposition and recognize that a lot of “sophisticated” potential counterparties are in a worse situation since they are paying other people to estimate a risk that even the “professionals” are predisposed to underestimate.
The point about being paid to mis-measure the risk is only one example of why asymmetric returns are so dangerous. Making the mistake outlined above is often profitable most of the time. The Pavlovian response to the repeated positive feedback is dangerous for two reasons. First, it is a tempting trade since most of the time it makes money. Entire businesses have been built on the returns. Second, the positive feedback reinforces the tendency to underestimate the risk. Thus, there is a tendency to “up the ante.” In businesses, this takes the form of increasing the risk exposure or even ignoring risk guidelines.
“Is this relevant to the average investor?” You bet it is! People often make exactly this kind of trade. This blog has mentioned selling naked puts before, but a far more common example is not having any “rainy day fund.” As absurd as it is, we have to force unemployment insurance. Unemployment, at least once in a career, is almost inevitable. Beyond that, not insuring against risks one can’t bare is almost too common to deserve mentioning. We’re debating whether people should have to carry medical insurance and have liability insurance when they drive. Sure, most of the time one gets away without either; thus avoiding having to pay the premium. It is continuous positive feedback. But, it only takes once to wipeout any premiums previously saved.
More in the investment area as we usually define it; think about why we have margin limits. Enough people get the likelihood of unusually large moves so wrong that it makes margin limits advisable as protection for individual investors and to protect the clearing houses.
However, my favorite example is liquidity risk. Investors get it so, so wrong. To illustrate, everyone reading this posting probably either has experienced or will experience times when most of their assets are illiquid. We aren’t just talking overnight or the mutual fund industry’s practice of redeeming at closing net asset values. Exchanges get closed down, markets dry up, and these things happen fast.
The probability of more routine fluctuations, even big ones like the recent events, are underestimated. They not only happen, but they last longer than most people realize. To illustrate, think about mark-to-market accounting for a minute. Aside from the absurdity of assuming the market is always right, we’ve almost enshrined the idea of continuous liquidity into accounting.
Even the probability of lesser forms of illiquidity get mis-estimated. Consider big, instantaneous changes in price (“discontinuities” or “gapping” in investor jargon). If one has investments, there is a good chance one asset experienced a gap in price during the time it takes to read this posting. It might be small, but it isn’t unusual. Some investment advisors recommend always using limit orders as protection against discontinuities being used to the investors detriment.
A disclosure seems appropriate. I have a strong preference for avoiding asymmetric payouts totally, and when they are unavoidable, I lay the risk off with insurance.
2) Complexity as a risk
Beyond asymmetric risk issues, the quote also implies a few other mistakes to avoid. First, my reading is that complexity is its own risk We’re not talking the complexity of the instruments being traded. Seriously, the parties involved all understood the instruments being traded. They weren’t that complex to people familiar with bond markets. The complexities that probably tripped them up were the complexity of their own organizations and the complexity of the RISK embedded in the instruments.
One could argue that complexity of the risk embedded in an instrument is complexity of the instrument. So, that leaves organizational complexity. The key quote is: “Royal Bank's huge loss raises questions about whether the ABN Amro managers knew what they were getting into…” That can be interpreted two ways: as either not knowing the risks implied by the guarantee or not knowing what the risk implied by the guarantee did to the Royal Bank’s overall risk. The quote addresses only the first. That’s possible, but seems less likely than the second.
But, the reader might ask: “How does that relate to an investor?” First, a direct implication is worth noting. If the second is true, it raises questions about the viability of the Basel II framework. It implies we are underestimating the level of risk in large banks operating under Basel II. Basel II could be conceptually correct, but impossible to implement in practice. There can’t be any doubt that the concept of hedging risk is a risk reduction strategy, but it is equally obvious that the more complex the risks being hedged, the greater the risk of gross mis-measurement somewhere in the process.
This reasoning would imply that growth as a risk reduction strategy has some inherent limitations. Further, it would seem logical that growth through acquisition, especially acquisition in new financial service areas, would carry the greatest risk. This could explain why banks’ initial acquisitions in non-banking areas often don’t work out. It wouldn’t be the only reason, but it probably contributes.
That would still limit the implications to investments in the financial service industry. However, if one views portfolio complexity as analogous to organizational complexity, it suggests additional implications.
First, if big financial firms with all their staff have trouble managing complexity, most investors are going to have a harder time. One should try to simplify away any complexity that one can’t clearly justify. Put, differently, know why each holding fits into the portfolio, not just why it is a good standalone investment. Balance the two: portfolio fit and standalone appeal. If an investment looks good as a standalone, but, for example, over-weights an asset class, use forced choice. Some asset in that asset class should be sold. A disclosure is appropriate because this discipline has worked so well for me that I might be overlooking limitations. A true fundamentalist would argue a total from-the-ground-up approach with each investment ONLY assessed as a standalone is better.
Second, financial firms can’t get hedges perfect: one should never assume a hedge is quantitatively right even if it is checked regularly. The quantitative values of hedges fluctuate just like the value of other assets. In other words, hedges have to be rebalanced periodically, just like any other asset in a portfolio. But, rebalancing isn’t enough. Hedges, just like any other asset, are subject to tail-risk. It even seems the tail-risks are greater with hedges than net long or net short positions. Perhaps it’s because hedges involve at least two positions. From what has been said, it should be apparent hedges don’t provide a perfect substitute for liquidity and don’t necessarily justify greater leverage risk. My disclosure is that I use traditional hedges sparingly, and always to hedge only one very specific risk associated with another position, not the position itself.
Third, when taking a position in a new asset, think of it as analogous to a corporate acquisition. If it’s new in type, expect errors. Some advisors recommend starting small; others admonish against any experimenting with a new investment style (i.e., “stick to your style”). Some advisors suggest doing it on paper first. My tendency is to only take on one new type of asset or type of trade at a time and to assume I’ll make some mistakes. My motto is “Stick to your style to make money. Experiment to learn.”
3) Know the difference between fees and profit
Perhaps it is how the transactions are reported, but one is definitely left with the impression that the pursuit of the fees that they would earn tempted these firms into trades that resulted in some losing investments. One might conclude that the point is irrelevant since fees are what financial firms are about; it’s their business. However, the point has relevance to non-institutional investors.
First, one should always remember that hedging away all risk is desirable in the financial service industry where there are fees to be made on both sides of the transaction. Most investors don’t enjoy that benefit. For most investors the fees flow out, not in. So, trying to do what the “big boys” do can be foolish. Risk isn’t something to avoid. It is inevitable for investor.
Second, fees can’t guarantee you’re not trading against your counterparty. You’re always trading against your counterparty. It is amazing that some people want to vilify GS for trading against “clients” they sold assets to. They don’t seem to realize that selling or buying an asset implies trading against one’s counterparty.
Third, fees appear to have encouraged participants to take on more risk than they could manage. For investors it is a reminder that fees are a consideration, but not THE primary consideration. Avoiding fees shouldn’t drive the investor away from a good investment any more than the pursuit of fees justified losing position by parties to the trades involved in this situation.
4) Information disparities
This is the heart of the civil suit against GS. So, it would be premature to go too far in discussing it until the suit is determined. However, there are some anomalies worth noting. One contention is that two facts weren't disclosed: (1) a short was the counterparty and (2) GS’s position. That these are an issue seems curious. They are information that one doesn’t have on any trade executed on an exchange. The anomaly is that regulatory reform would expand the role of exchanges in derivatives trading, thus extending the absence of this information to more of the derivatives market.
In the March 4th posting, The Hedged Economist argued: “Bringing more derivatives (i.e., like some standard interest rate and default swaps, some commodity hedges, etc.) onto exchanges makes sense.” Both the use of an exchange and a clearing house were endorsed. If that position is correct, one might conclude that the information at issue in the Goldman Sachs civil suit isn’t “material.” Yet, the April 23rd posting noted investors who successfully traded the same instruments “put effort into understanding whether their immediate trade was with a counterparty or a middleman.” It goes on to suggest that it’s a good idea for investors to do the same. That would suggest that the information is “material.”
The April 23rd posting, when discussing the successful traders, went on to say “Once the people making the trade understood who their counterparty would be, their primary concern wasn’t motive. Their concern was solvency.” Their concern wasn’t whether their counterparty thought the price would go up or down. They understood that their counterparty probably had an opinion different from theirs.
But interestingly, a second issue related to the suit is the fact that a short seller was involved in structuring the asset. The irony here is that a short is often the originator of many positions in any asset without it being disclosed. It’s very common in options. But even with stocks, short sellers may be selling an investor a stock or selling it to an individual’s mutual fund. Most successful investors have no reservations about buying without knowing whether a counterparty is a short seller.
Ultimately asymmetric information is a fact of life. The court will decide what information should have been disclosed. Putting limits on the information asymmetry is desirable. But, from an investor’s perspective, making assumptions about the direction of the asymmetry can be as big a mistake as assuming there is no asymmetry. Furthermore, and far more importantly, it is essential for investors to decide what information is important to their own investment decisions. The courts can’t do that for them.
It would be unfortunate if GS’s role or Paulson’s success keeps investors from looking beyond the superficial media coverage. This could be an instance where we aren’t left with only our mistakes as examples of what doesn’t work. The most obvious lesson for investors is how dangerous it is to just blame GS if the recent crisis hurt one’s portfolio. Many people won’t bother to look at how they managed their liquidity or their leverage levels, or examine the risks associated with their investment strategy. While it’s important to learn from others’ mistakes, it’s essential to think about how they compare to one’s own.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment