Bulls and Bears: How Asset Prices Evolve

In last week’s post I mentioned three stages in the evolution of a market:

Identifying full-blown bubbles is easy. What’s not so easy is identifying the transitions that bookend a bubble. It’s not easy to know precisely when a rational, fundamentals-driven boom will morph into an irrational, sell-to-the-greater-fool frenzy. It’s not easy to know precisely when an irrational frenzy will reverse into an equally irrational stampede for the exits.

These three stages – rational boom, frenzied bubble, irrational panic – are in fact just three out of a total of six stages in my own idiosyncratic (and highly unscientific) taxonomy of bull and bear markets. Here’s how it works.

The first stage in any bull market is what I like to call the bounce. A sector or asset class that has been moribund for years or even decades suddenly starts rising in price. This could be due to exogenous shocks such as regulatory or technological changes; it could be due to Schumpeterian creative destruction, wherein prolonged low prices have driven out the weak and created a breeding ground for strong innovative companies; it could be due to simple cycles in supply and demand like the ‘commodity supercycle’. Whatever the reason – and often the operative reasons are not evident till many years later – prices begin to move upward. This is the bounce.

Typically during the bounce stage prices increase but the asset class remains unfashionable; only a few visionary investors recognize the bounce for what it truly is, the harbinger of a prolonged bull. Above all people don’t recognize the reasons for the bounce. Indeed, proselytizing for an asset class or sector during its bounce phase is a thankless job; you will probably get sniggered at by television talking heads for your trouble.

The second stage in a bull market is what I like to call the boom. During this stage, price rises have begun to attract more attention from the investment community. This is a stage of diffusion: the investment story spreads beyond its first few evangelists to an ever-increasing audience of relatively well-informed investors. To a large extent the strength of the sector becomes conventional wisdom. But prices continue to rise; it is not that contrarianism (going against the conventional wisdom) has failed; it is merely that the fundamentals continue to be so strong that they outweigh any technical factors.

The third and last stage in bull market is what I like to call the bubble. In this stage, the fundamentals have ceased to matter. In fact, the growth of the boom years has created sufficient supply to cause fundamentals to tilt to the opposite direction. But nobody notices. Drawn by strong performance, ever more investors flood into the sector. High prices create their own self-reinforcing dynamic. Positive feedback, mass self-delusion, ‘this time it’s different’, new paradigm stories, ‘permanently higher plateaus’, huge quantities of supply, sectoral employment shifts, dodgy startups, reality TV shows, easy funding – these are all symptoms of a bubble phase.

It’s pretty easy to distinguish between the three stages of a bull market. Certainly nobody could mistake a bounce stage (in an obscure and unfashionable sector) for a boom stage (where the sector is widely known for its strong fundamentals, albeit less widely invested in). Still less could anybody mistake a boom for a bubble: in a boom the fundamentals still rule, in a bubble fundamentals have gone out the window and the greater-fool theory rules. (Though well-meaning but misguided analysts inevitably try to justify bubble-era prices and try to coax the market into some sort of fundamental-based story; this usually involves invoking a new type of fundamental).

Just as a bull market has three stages, so too a bear market. The three stages of a bear are fairly accurate mirror images of their bull correlates.

First comes the blowup, in which the excesses of the bubble are purged. This purge is often quite dramatic, as the positive feedback loop that fueled the expansion reverses direction, causing prices to fall as precipitously as they previously rose. The excess liquidity that helped inflate the bubble is withdrawn with quite astonishing rapidity, leading directly to various closely related phenomena that are emblematic of a panic: the flight-to-quality reflex, the cash-is-king psychology, and the dynamic of the liquidity-death-spiral.

The next stage in the bear market is the bust. This is a long drawn out decline in prices as the market works out its overhang of excess supply (created in the boom) and anemic demand. The bust can last for years or (if markets are not allowed to clear) even decades.

The final stage of the bear market is the bottom. This is not a single point but a very lengthy period in which investor interest wanes, volumes and volatility decline, and sector news gets relegated to the inside pages of the financial dailies. Of course the bottom merely sets the scene for the next stage in the market, the bounce of the next bull market. And thus the circle is complete.

History: It Ain't Just Bunk

Human beings are good at interpolation, passable at extrapolation, bad at identifying inflexion points, and downright terrible at processing one-off events. It’s no coincidence that these skills are, sequentially, associated with increasing investment success: the harder it is to do something, the more money one makes for doing it.

This particular progression from easy to difficult is not merely the artifact of some deep-seated behavioral tendency or evolutionary bias. Deterministic and presumably unbiased algorithms, faced with unprecedented events, perform just as badly as humans. This is only to be expected: the very word ‘unprecedented’ implies that there is no baseline to build from or compare with, a circumstance under which most algorithmic approaches tend to flounder.

Unfortunately for all concerned, real life is full of one-off events. What we call history is, as Rudge memorably puts it, just one bloody thing after another. And that’s precisely why I’m suspicious of attempts to mindlessly trawl through past data for aggregate patterns. Every episode is different; every episode is new.

This does not mean that history should be discounted entirely. Quite the contrary. A deep and broad knowledge of history (and not just the history of the markets!) is essential to becoming a successful trader. Events may not repeat themselves exactly, but they certainly rhyme; the trick is to find out what they rhyme with.

So how does one accomplish this trick? Regular readers will know the answer: by asking ‘why’. Questions such as ‘what’ or ‘when’ or ‘which’ or ‘how much’ are no doubt useful when it comes to short-term, tactical trading, but they are limited in their ability to throw light on long-term, strategic trends. Asking ‘why’ a particular historical event turned out the way it did, on the other hand, is the first step towards recognizing its kinship (or lack thereof!) with seemingly similar events developing today. Understanding the past is the key to understanding the present, to say nothing of predicting the future.

This sounds overly abstract but in truth it is anything but; the technique of asking ‘why’ at all times can (and should!) be used to analyze not just big-picture historical movements, but also individual trades. Indeed, finding out why a particular trade worked while others failed is a key component of the trader’s art.

Here’s an example from my own career trading bonds. My portfolio was, in general, designed to capture or monetize excessively rich risk premiums (curve, liquidity, capital structure, you name it). Risk premiums of course tend to widen in times of market stress, so my portfolio behaved as if it were short event risk. To hedge against this I invariably had a long position in Fed Funds futures and the first few Eurodollar contracts, confident in the knowledge that any ‘flight-to-quality’ would send these assets higher. (Also, in truly extreme cases the Fed could be counted on to step in and cut rates, helping the front of the yield curve.)

I was not alone in this practice. Here’s an excerpt from an interview with Christian Siva-Jothy, former head of prop trading at Goldman Sachs:

Being long fixed income is like a synthetic long gamma trade. More than 90 per cent of the time, if there is a major dislocation to the economy, fixed income will rally. I sleep better at night doing that.

Insurance is not the only reason to be long bonds. There’s also the widely-held belief that rallies tend to be slow grinding affairs while selloffs tend to be sudden sharp shocks¹; it’s a lot easier to ride the former than it is to time the latter. Here’s Siva-Jothy again:

Bear markets in fixed income are very short with powerful rallies. You can make money during a bear market but you have to time your trades perfectly.

As a matter of fact, most successful bond traders of the recent past, like Siva-Jothy, have had a perpetual long bias, and have justified it on similar grounds.

Looking back though, I wonder if this is not just post facto rationalization. After all, the Treasury market has been rallying more ore less continuously for the last quarter of a century; long bond yields have gone from 15.5% in 1981 to 2.5% in 2008. You would have had to be a spectacularly incompetent long-biased trader not to make money over this period. Conversely, no matter how good you were at trading from the short side, you’d have been hard pressed to make big returns in such a strong bull market. And that’s why most bond traders, through experience or by selection, tend to have a bullish stance².

All very well, but so what? So this: what if bonds turn? What if the 30-year bull market was a one-off event that will not be repeated, rather than a trend that will continue³? What if 2008 marked the low in bond yields? What if rates stay steady or trend higher over the next decade or two? Will the front of the yield curve still serve as an event hedge? Will rallies continue to be protracted and selloffs continue to be compressed? Right now, nobody knows for sure, but these are questions worth keeping in mind. A trader who does otherwise – who merely trades from the long side without asking why being long Treasuries worked in the past – risks being blindsided.

Footnotes

# ¹Here’s a cherry-picked illustration:



# ²Of course, this explanation merely pushes the question back one level. Why did the bond market rally for 25 years? That’s a question that deserves a full-length post in answer.

# ³30 years is, admittedly, a long timeframe for a ‘one-off’ event, but note that the current bull market was preceded by the greatest bear market in US Treasury history. Perhaps the entire rally in rates since 1981 is merely reversion to the long-run (pre-bear) mean.

Feedback in Financial Markets

In a previous post, I mentioned that bubbles were characterized by – indeed, defined by – positive feedback. This idea, and more generally, the importance of feedback in driving market dynamics, deserves a lot more ink. Here’s a first installment.

Classical economics is often concerned with analyzing various equilibrium outcomes (“comparative statics”). These outcomes are usually generated or maintained by some sort of negative feedback. The simplest example is that of security prices. Under the efficient markets hypothesis, each security has a fair price reflecting its ‘fundamental value’; furthermore, this fundamental value is known to market participants in aggregate. If the actual market price drops below this value, people step in to buy the security; if the price rises above it, people step in to sell. As a result of this negative feedback, the market price equilibriates to its natural or fundamental value.

Unfortunately markets do not always tend to equilibrium. Negative feedback is not always the dominant mechanism at work. And fundamental value is not always well defined. Bubbles provide a clear example of each of these counterfactuals.

In a bubble, the dominant mechanism is positive feedback; the key to understanding bubbles is understanding this positive feedback. How, then, does positive feedback arise?

The most obvious explanation is the conventional one: positive feedback is a consequence of irrationality in the market. And there’s certainly an element of truth in this explanation. Greed, self-delusion, unjustified extrapolation, caring more about relative returns than absolute profits (a.k.a. “keeping up with the Joneses”), conformism (a.k.a. “if everybody else is doing it why can’t we?”), confusing the improbable with the impossible (“house prices will never go down nation-wide”) and other persistent behavioral flaws lead inevitably to bubbles. This has been true throughout the history of speculation.

But one doesn’t have to invoke irrationality to explain positive feedback. Positive feedback can arise quite naturally when rational traders encounter flawed institutional mechanisms, as my previous post makes clear. Short-term incentives, asymmetric outcomes, incomplete information and firm-wide pay structures could all lead to perfectly rational actors taking actions which lead to positive feedback and hence bubbles.

Both of these are what I would call ‘technical’ explanations, in that they depend on trader behavior (which has various causes) to move market prices away from some underlying fundamental value. But there is another, and to my mind more interesting, form of positive feedback in which the fundamental values themselves change.

Consider this example from the FX markets. A currency strengthens. This acts like a tightening of monetary policy. Hence inflation expectations diminish. Hence the currency strengthens further. The initial move has thus changed the underlying fundamentals so as to justify itself; it has become a self-fulfilling prophecy.

Or consider this example from the equity markets (private email from my friend WB):

Markets prices impact fundamentals. If Amazon's stock price goes up as it did in 1999-2000, it makes it that much easier to raise capital either from debt markets or from equity markets. If Amazon raises more money it can invest more and make improvements which make the future look that much brighter. That pushes up prices even higher. That's not a negative feedback cycle. In fact it's downright positive feedback. This can go on for a very long time, but then one day the reality just doesn't offer as much as was priced in and we have an enormous collapse which again acts in a positive feedback way. So in the end over medium horizons, markets can be mean-averting or create trends while in the longer term picture they are mean-reverting.

Or, closer to home, consider this example from the real estate markets (lifted from Wikipedia):

Lenders began to make more money available to more people in the 1990s to buy houses. More people bought houses with this larger amount of money, thus increasing the prices of these houses. Lenders looked at their balance sheets which not only showed that they had made more loans, but that their equity backing the loans—the value of the houses, had gone up (because more money was chasing the same amount of housing, relatively). Thus they lent out more money because their balance sheets looked good, and prices went up more, and they lent more.

Of course, the conditions required to foster a ‘fundamental’ positive feedback loop don’t arise very often, but when they do, the outcome is dramatic.

The final word belongs to George Soros, who treats feedback as a special case of his larger socio-economic theory, ‘reflexivity’. Soros’ book The Alchemy of Finance contains many more examples of ‘fundamental bubbles’; rather than quote them all, I’ll leave you with two short excerpts from this 1994 speech:

I must state at the outset that I am in fundamental disagreement with the prevailing wisdom. The generally accepted theory is that financial markets tend towards equilibrium, and on the whole, discount the future correctly. I operate using a different theory, according to which financial markets cannot possibly discount the future correctly because they do not merely discount the future; they help to shape it. In certain circumstances, financial markets can affect the so-called fundamentals which they are supposed to reflect. When that happens, markets enter into a state of dynamic disequilibrium and behave quite differently from what would be considered normal by the theory of efficient markets. Such boom/bust sequences do not arise very often, but when they do, they can be very disruptive, exactly because they affect the fundamentals of the economy

...

For instance, in a freely-fluctuating currency market, a change in exchange rates has the capacity to affect the so-called fundamentals which are supposed to determine exchange rates, such as the rate of inflation in the countries concerned; so that any divergence from a theoretical equilibrium has the capacity to validate itself. This self-validating capacity encourages trend-following speculation, and trend-following speculation generates divergences from whatever may be considered the theoretical equilibrium. The circular reasoning is complete. The outcome is that freely-fluctuating currency markets tend to produce excessive fluctuations and trend-following speculation tends to be justified.

And there you have it, straight from the greatest trader of the twentieth century. Further comment would be superfluous.

Bubbles and Scale Invariance

In yesterday’s post I mentioned that bubbles were exponential, scale-invariant and self-similar, making it virtually impossible to time their collapse.

Let’s flesh out this assertion by looking at a particular market index.

For the first 17 years of its existence, this index had a mean of 100 and a standard deviation of 56. (Prices have been scaled to avoid easy recognition). That’s a pretty stable time series.

Then something happened. Over the next 8.5 years, the index went from a starting value of 200 (already near the upper end of its previous range) to a value of 700. What’s more, this rise took on exponential, maybe even super-exponential characteristics, as the graph below makes clear.


Would you sell? If you did, you’d be out of luck. Because over the next 25 months, the index went from 700 to 1100. Once again the rise looked exponential or better:


(Note that this graph has the same start date as the previous one, but different scales on each axis).

Would you sell? If you did, you’d be out of luck again. Because over the next 15 months the index went from 1100 to 1500, with the pace of expansion growing ever higher


(Once again, this graph has the same start date as the previous two, but different scales on each axis).

Now would you sell? How much further and faster can the market rise? The answer is, quite a bit. Over the next 5 months the index rocketed from 1500 to 2800. If you had sold the index at any of the previous junctures – and note that at each of those points, the graph looked convincingly bubbly – you would almost certainly have been carried out at a loss.

2800 was, in fact, the high; over the next 31 months the index dropped all the way back to 600. Here’s the full graph, with dates and true (unscaled) values.


I’ve marked the extrema of each of the previous graphs onto the composite graph, to demonstrate how scale-invariance works. Although zooming in on any sub-graph gives the impression that it’s an exponential curve about to pop, these curves just get lost in the main graph. It’s not easy to time bubbles.

Postscript: The index in question is of course the Nasdaq composite in the days of the dot-com expansion. Interestingly, Alan Greenspan warned about ‘irrational exuberance’ in December 2006, shortly after the first of the graphs above. Three years later he had changed his tune (‘capitulated’?) quite substantially.

Implicit Regulatory Arbitrage: The Puts-Payers Trade

Yesterday’s post revealed how (and why) a large portion of the financial industry’s revenues came to depend on explicit regulatory arbitrage. This is fairly common knowledge, and should come as no surprise to industry observers.

What’s not so well known is that many ‘classic’ arbitrages, which appear at first glance to be regulation-independent, also depend implicitly on regulatory asymmetries to work. The textbook example is bond futures arbitrage. While anyone can buy bonds, some market participants are forbidden to sell bonds short. To express a bearish view, this latter group has to sell bond futures. This makes bond futures systematically cheap relative to cash bonds. Arbitrageurs have only to take the opposite side of this transaction to make easy money.

Of course, the classic bond futures arbitrage no longer exists (‘”too many eyeballs”), but other, subtler examples abound. Consider a trade that was very popular with fixed income arbitrageurs earlier this decade: the puts-payers combo. This trade involves selling Treasury puts and using the proceeds to buy payer swaptions, for zero net premium. Both the puts and the payers are struck slightly out of the money.

How does the trade work? If the market rallies or stays rangebound, the options expire worthless. But if the market sells off, the options are exercised, and the trader finds himself long Treasuries and paying fixed in swaps – in other words, long swap spreads. So, the trader is making the bet that ‘swap spreads will widen in a selloff’ – and he’s making this bet for free (remember, there’s zero net premium to enter this trade).

Is this a good bet to make? Let’s look at some history:


That’s a pretty strong relationship between two supposedly independent variables, and hints at some serious inefficiency in the bond market. What’s going on?

The answer is simple. Just as in the bond futures trade example described above, the arbitrageur in the puts-payers trade is taking the other side from entities who are forced by regulations to behave sub-optimally. In this case, these entities are the government sponsored agencies Fannie Mae and Freddie Mac.

Some background may be useful here. In the early years of this decade, Fannie Mae and Freddie Mac were massive players in the bond market. At the time, they had very large mortgage portfolios which were characterized by significant ‘negative convexity’. This characteristic meant that when the market rallied, they needed to buy; and when the market sold off, they needed to sell, in order to keep their portfolios properly hedged.

Now, Fannie and Freddie, being government agencies, faced restrictions on their size and trading activity. Consequently, they decided to do the bulk of their convexity hedging (described above) in swaps rather than in bonds, since swaps are off-balance sheet instruments, while bonds have to be reported. Every time the market sold off, Fannie and Freddie would be out there selling (paying) in swaps, in size. Swaps would therefore underperform bonds in selloffs; hence swap spreads would widen. (Note that the mortgage market is much larger than the government bond market; hence Fannie’s and Freddie’s trading actions, determined by the former, would invariably move prices in the latter.)

This pattern repeated for years and years: it was that rarest of beasts, a persistent and captureable anomaly in the market. Many arbitrageurs took advantage of this, via structures like the puts-payers combo.

Did these arbitrageurs generate alpha? Well, yes and no. Within the context of the bond market, the answer is yes: the agencies behaved sub-optimally, and thus transferred wealth to the arbs. But viewed at a larger scale, the answer is no: the agencies behaved rationally by paying the arbs to move their interest rate exposure off-balance-sheet. The arbs were therefore being compensated for a service they were providing; they were harvesting alternative beta rather than capturing alpha.

(This is just a particular case of the general truth that any alpha is merely a beta within a larger universe. In this case, bond-market alpha turns out to be service beta. I will return to this concept in future posts.)

Of course, even this persistent anomaly couldn’t last forever. The previous scatterplot shows data from April 2000 through March 2006; the following one shows data from April 2006 through August 2008 (Fannie and Freddie went into conservatorship in September 2008):


The inefficiency has all but disappeared. What happened?

The obvious answer is the correct one: over the years, Fannie and Freddie scaled back their interest rate trading activity considerably. Fannie Mae, for instance, shrank its mortgage portfolio (the “owned balance sheet”) from 917 billion in 2004 to 728 billion in 2007. Over roughly the same period, Fannie reduced its duration gap (a measure of the mismatch between assets and liabilities, and a strong proxy for the portfolio’s negative convexity) from over 1 year to less than 1 month, by buying swaptions and issuing callable debt. With a significantly smaller and better-hedged portfolio, Fannie simply didn’t have to trade that actively.

Obvious, yes, but only in hindsight. An arbitrageur who tried to play the puts-payers game after 2006 would not have made any money. Once the regulations changed (and make no mistake, Fannie and Freddie’s portfolio redesign owed a great deal to regulatory pressure), the regulatory arbitrage disappeared.

Why does any of this matter? The case study of Fannie Mae, Freddie Mac and the puts-payers trade highlights a theme that I will return to again and again on this blog: the importance of understanding the source of your returns. It’s not enough to spot an inefficiency (opportunity) in the market; you must also know why the inefficiency exists. Only then can you avoid being blindsided when circumstances change and the opportunity disappears.