What is about math that makes us feel comforted? Numbers are objective, of course, but the using of numbers is not. Even in the hard sciences calculations are not strictly calculations for their own sake, they are interpreted and therefore given subjective meaning. I don’t intend to detour this argument into a teleological one, but in some ways that just can’t be helped.

We know about statistics in econometrics, but its parallel and really in many ways the inspiration for the use of math in economics derived from its long-standing application in finance. For as long as humans have wondered they have tried to predict the future. Almost every system invented for that task has tried to do so by searching the past for recognizable patterns. It is here where algorithms and regressions may be so alluring.

But to get from that to a robust forecasting paradigm requires truly more computational skills than are available even today, let alone ten or twenty years ago. That hasn’t stopped people from trying, of course, and it likely never will. The Spanish once scoured the Americas for El Dorado, the mythical lost city of gold; to a modern shadow bank, untold riches aren’t found in some far flung rainforest, they are found instead on a balance sheet governed by the secret black box formula.

The first outfit believed to have found it was LTCM. Its rise and fall have largely been lost to history, surpassed in every way by what happened in 2008. But what went on ten years earlier told us everything we needed to know about the crisis, including the idea of liquidity in the mathematical, modern wholesale sense; the system that was driven by far more derivatives as money than money as money.

In September 1998, the Federal Open Market Committee of the Federal Reserve sat around the table to discuss the intricate problems created by the prospect of LTCM’s complete demise, at times marveling at its feats while also being terrified by what that might represent of a future far different than the arrangements they all knew and understood.

CHAIRMAN GREENSPAN. Somebody mentioned to me that Bankers Trust had an August balance sheet for LTCM. Is that true?

 

VICE CHAIRMAN MCDONOUGH. Yes, but the balance sheet is a relatively small piece of the whole action because so much of the latter is off-balance-sheet.

Traditional accounting systems are no match for this forward way of thinking. I don’t mean “forward” in the sense of LTCM representing progress, rather everything they did was modeled into the future, brought back into present values and then recorded from those. Obviously in such a situation there is a great emphasis on those calculations, which was, to be frank, LTCM’s game. They were the first to do it so extensively and so they had an enormous asymmetry of information, not the least of which was their roster of financial gods, those men who pioneered all this stuff.

That gave them a tremendous advantage, at least in the beginning.

MR. FISHER. It was something of a signature for this firm to insist that if a counterparty wanted to deal with them, there would be no initial margin. Not many other firms have gotten away with that.

 

MR. GRAMLICH. It goes to your bedazzlement effect.

The “bedazzlement effect” was a factor that Chairman Greenspan raised, namely that when LTCM showed up at whatever bulge bracket Wall Street firm, who were they to tell Bob Merton and Myron Scholes to put up margin? They had all the mathematical power on their side, so in the mid-1990’s LTCM amassed gigantic positions far beyond their traditional balance sheet and nobody really knew how or why; including LTCM, as it turns out.

The issue for me today is how to define liquidity in the sense of this kind of systemic arrangement. There is a funny quirk to it that I don’t believe is appreciated in the way it has played a major role in 2008, as well as 2011 and 2014-17. It has to do with the math of future expectations coupled with behavior that you would normally associate with a bank run. The FOMC members in September 1998 only briefly touched on it, and it is clear, at least to me having studied the transcripts from the intervening decades, they didn’t truly appreciate the significance of what they were seeing. It was largely confirmation bias, where these policymakers really thought this one firm was a boutique, an aberration soon to be forgotten in the timeless world of banking.

MR. FISHER. There is a fair amount of government credit in these assets, but there are a lot of other assets also. Swap agreements are their instrument of choice, and that is how they got to a $1.45 trillion off-balance-sheet position on August 31. By the time we were looking at that position during the weekend eight days ago, the firm clearly had lost a lot of capital. Other firms that looked at their position in greater detail than we were able to thought the off-balance-sheet had shrunk to around one trillion dollars by the third week of September. The balance sheet leverage ratio was 55 to 1 by the time we looked. The off-balance-sheet leverage was 100 to 1 or 200 to 1–I don’t know how to calculate it.

 

 

MR. SPILLENKOTHEN. Initially, these transactions in derivatives were in a sense unsecured. Once the exposure reached a certain point, then the lender, the bank, had the right to demand collateral to cover that. That is, when the mark-to-market position and the current exposure built up to a certain point, then the banks asked for the collateral. According to the numbers we have now, it appears that the mark-to-market positions of these big institutions are largely covered by collateral. So, it does not seem that there is much loss in their current mark-to-market positions and other derivatives positions. One of the problems is the potential future exposure. I am going to say this in English. If markets keep moving away from them in the wrong direction, their future exposure could be very large and they might not have the collateral at that point in time to cover that exposure. [emphasis added]

Losses are never really the problem; Lehman Brothers didn’t fail on losses, nor did AIG. In fact, AIG’s portfolios taken over by the Federal Reserve in late 2008 and put into the Maiden Lane hedge funds actually made money for the Fed over time. What killed those firms was modeled future exposure, the key component of which is volatility, meaning expected volatility. The way these processes are set up, expected VOL can have a tendency to become exploding VOL.

At Time Period 0 when volatility is normal, meaning within designed and expected tolerances, an LTCM or a Lehman Brothers can get their positions funded by pretty much anyone they like. In the case of LTCM, they just showed up and demanded X and was given X because nobody really understood what they hell they were talking about. In the case of Lehman a decade later, or anyone else, they had by then acquired the ability to speak the language of LTCM fluently and thus their math all agreed on these transactions. If Lehman wanted to do X but couldn’t get JP Morgan to take the other side because JPM’s math was different, then Lehman goes to Bear or Morgan Stanley or whomever else where the math speaks on the same terms.

But it doesn’t matter how dazzling you can be, or how impenetrable your calcuations, if things start to go wrong the sharks circle no matter what or who you are.

VICE CHAIRMAN MCDONOUGH. It was fascinating to me that every head of the 8 to 10 firms that I talked to subsequently during the day brought up the problems of LTCM/P independently of anything that I said. They gave particular emphasis to what in their view would be the very serious problems that a failure of that firm would create in financial markets. They were not talking particularly about the problems that such a failure would cause for their own firms but rather about the problems that it would cause for the financial markets in general.

Everyone on Wall Street in September 1998 was talking about LTCM not just because they were worried but because they smelled opportunity. Maybe Wall Street didn’t understand the intricacies of LTCM’s models but they didn’t need to in order to figure out when they weren’t working. Thus, when LTCM shows up desperate to get refunded in X or even just rolled over because they say VOL is still Y, you suddenly have the confidence to counter with Y + some additional premium; not because LTCM, or you, is at risk today, but because you can make the math sing about tomorrow to your tune rather than theirs. And you further know that if you demand a premium based on these new calculations, so will everybody else. Information asymmetry can be the greatest thing ever, at least until it becomes symmetry.

A firm that finds itself in this situation is not the object of mercy, but a target to be ripped apart no matter the systemic consequences. A Bloomberg article published today revisiting Lehman’s failure from the perspective of “giddy” Citigroup traders touches somewhat on this anti-liquidity process.  It gives us a small but fairly open sense of it given the original source material (internal Citi communications provided as part of a lawsuit presented by Lehman creditors).

“Ringing the register, homey,” Thomas Giardi, a trader in the bank’s credit derivatives trading unit, said in a chat message at 6:40 a.m. on Sept. 17, two days after Lehman’s bankruptcy. Subject line: “YOU MAKING $$$.”

 

 

“This is the time to make a lot of money,” Carey Lathrop, head of Citigroup’s credit division, said in a Sept. 17, 2008, recording, according to Lehman’s exhibits. “There are a lot less competitors… and there’s a lot of bid/offer spread. And people have to do things.”

That last part is what really matters; “and people have to do things.” When the math goes horribly wrong, desperation comes in the form of exploding future liabilities that have to be dealt with in the present no matter how far into the future they may be. In that situation, the person on the other end of the phone has all the leverage, literally in these cases, provided that they are not swept up in the hurricane, too. Lehman had failed but there was still business to do. How do we define liquidity here?

It isn’t extortion because Citigroup was not threatening to do what Lehman hadn’t already done to itself; but it is extortive beyond Lehman. The most recent (likely) example is what I wrote about yesterday, the case of Japanese banks locked into a spiraling cost structure obtaining “dollars”, largely FX, in order to redistribute them, I believe, to China and elsewhere around Asia. Once the math turned against them, modeled VOL due to Chinese exposure (as China’s economy “unexpectedly” stopped growing because the global economy “unexpectedly” did too) and then JPY that only kept going the “wrong” way, without a lifeline they were easy pickings for the always circling sharks – until the BoJ extended them a lifeline and, with a little luck, a possible (or possibly brief) way out.

Though we talk of liquidity and even might call it a run, there isn’t a dollar anywhere to be found; not a single Federal Reserve Note or bank reserve to be traded anywhere back and forth. Worse, most of what happens in these kinds of disasters like LTCM, Lehman, and the “rising dollar” doesn’t ever see the light of day on a balance sheet. It happens, often quickly, off in the far reaches of these virtual domains. The Federal Reserve does not rule here, it barely even rates. As confused as Greenspan et al were in 1998, two decades on those who have followed still are. They thought LTCM was an anomaly when its math about the future was the future.

What does rule is the mathematics, but it is not the math of science or objectivity. It is the 21st century math of money, and we would do well in whatever comes after this to, as best as we may be able, kill it and start over without it.