Monday, March 17, 2008

Feedback Trading's Hollow Victory

My friend, and part-time drinking companion, Macro-Man, whom I respect much, flippantly employed the title late last week of "Will the last one to leave please turn out the lights?" in regards to the spate of Hedge Fund blow-ups. I say flippant because I daresay this doesn't reflect the actual situation. For every Drake, Carlyle, Sowood, Amaranth, Peloton, Go, and yes and of course merely brow-beaten like SocGen, there is a Winton, Citadel, Brevan, Ospraie, Caxton, SAC or Tudor, or increasingly a bevvy of non-fundamental systematic investors popping cork on the vintage stuff as they do their victory lap. Not to say that there is a winner for every loser - for this is fallacy: there isn't. But you see, I am fascinated by the structure and nature of market participants - particular the more active traders who, in frenzied turnover, increasingly make up a greater percentage of daily value traded. And cannot help wonder whether the winners and losers are merely "the right" and "the wrong", respectively, or whether market Darwinism is not in the process of yielding something more pernicious that a winnowing of the most feeble investors and less-adept game-players.

Imagine for a moment that the winners and losers have, in addition to being right or terminally wrong, archetypical investment styles associated with them. Imagine that the future health of the market (at any point in time) is dependent upon the continued representation of diverse investor phenotypes. Now go one step further and imagine that Darwinian natural selection has wiped out an inordinate amount of the contrarians, be they the value-players, real short-term liquidity-providers (e.g. jobbers, market-makers, block-desks, independent floor traders or once-successful contrarians like Alliance Cap;s Alfred Harrison), en masse representing substantial risk capital.

Is this problematical? Perhaps not if the successful, i.e. those that were "right", were randomly victorious, OR if they were "right" on the basis of idiosyncratic but ephemerally-correct fundamental analysis that, next time around, would amongst the prior winners, yield equal amount of losers. But, as a thought-experiment, imagine what might happen if they [the winners] were predominantly of a single species, or reasonably correlated investment style or approach, be they trend-followers (CTAs), or momentum-oriented, systematic-macro employing price-based quant models for short-term predictive purposes, or order-sniffing micro-structure strategies - all which share traits that arrive at similar positions, for similar (essentially non-fundamental) reasons. What would this mean for markets, price behaviour and volatility, and the relationship between shorter-term prices and their more fundamental longer-term equilibrium?

Getting to the point, might a meaningful shift in the balance of "styles" pursued have a dramatic impact upon behaviour, elevating volatility, diminishing liquidity (in the market sense of size at a price), and elongating trends and their departures from longer-term fundamental equilibrium prices where real supply meets real non-speculative demand?? More to the point, hasn't this been happening in a almost continual process of natural selection over the past six years, accelerating recently with the rapidly bifurcating fortunes of all manner of traders, investment managers and hedge funds?? What is the practical result of a market with a paucity of liquidity providers and more value-oriented investors willing to warehouse risk, contra the prevailing trends? Alan Brown, Investment Head atop $280bn at Schroders in London said o Bloomberg:
I am struck by two thoughts. There are an awful lot of assets out there that are offering very attractive terms. On the other hand, we have an entirely dysfunctional market at the moment where marked prices are notional and if you want to trade, the bid falls away from you dramatically..."

He means liquidity withdrawn due to "The Credit Crunch", but the same dynamic might result from a rapidly changing balance in participant makeup where those with feedback-oriented strategies rule the roost, and amplifying the smallest of signals, reinforcing oft-divergent trends from the underlying market, and in the process disintermediating real buyers and real sellers, raising price volatility and uncertainty, both negatives for markets and economic efficiency.

Now we might also sensibly ask whether the survival of the fittest might produce better economic outcomes, for ridding markets of feeble should be an unmitigated good, no? To answer this, we need to ask "why?" this has come about at all, i.e. why has feedback trading has trounced human judgment? . The "why" has probably been globalization, poor policy extremes in US (fiscal, monetary and energy) & China (international and domestic monetary and trade) , and GCC, as well as the lack of policy agreement and coordination amongst nations. As for whether such natural selection is "good", I have a suspicion it is not. While such mimetic and feedback-oriented strategies still rely upon "the signal" from somehwere, what if at the source, "the signal" or impulse is manipulative or predatory (remember Amaranth?) in origin, or worse sheer fad or folly (remember ? In reality, most commodity markets and exchange contracts are small in relation to the financial economy, and so are ripe for manipulation and abuse by large, clever, or large and clever particpants' size and relative market power. If the sources determining market price and initial impulse were on the other hand real supply and demand, then perhaps the amplification of such allocative outcomes would serve to create signals and move prices towards longer-term equilibrium faster, thereby promoting efficiency. But where, for example, in Crude Oil, Wilbur Ross proffered on Thursday, more than 30% of trade is pure financial leveraged spec, much which is the sub-category type described here, one would have good reason to view their presence in regards to market efficiency, as, in the words of Donald Rumsfeld, "rather unhelpful" for a speculative collapse might, yet again , send alternative back in to the realm of the less-than-economic faster than a less-high but more stable price regime might accomplish.

Finally, one might ask the logical question: what sort of nightmarish doog-chasing-his-own-tail market regime follows such outcomes when the feedback traders - now dominant - are destined to trade amongst and against themselves in a nightmarish hell (for systematic traders) of failed breaks, large reversal gaps, that is perhaps the inevitable rent for the extended runs of the present, after which ummm errrI guess Citadel will own The World??!?!


kristiina said...

Ahhh, mmm... such an excellent post (again).

So, the thugs have taken over the watering-hole. What should the thirsty do? Hard to defeat the thugs. Probably must find/create/conjure a new watering-hole. It will take a little time for thugs to find it, especially if they are in a nasty fight over who gets to rule the old watering-hole.

"Cassandra" said...

I have some opinions in regards to feedback tracking and traders, but for this post, I hope I didn't use them in the pejorative. The thugs are people who push markets at public expense for parochial gain. Maybe, like the mythical but archetypical Col Nathan Jessup, they/it is a necessary evil. But to arrive at that conculsion, one should play around with the thought experiment as I've done - probably in a mathematical hypothetical laboratory market environment.

Macro Man said...

To quote Yogi Berra, it's like deja vu all over again!

Surely the same could have been written about the "fundamental" quant-based strategies, epitomized by Global Alpha, that were phenomenally successful (indeed, one might argue, hegemonic in certain spheres)...until they weren't.

Trend followers will presumably meet the same type of end, and while they are indeed a frustrating leader of the pack, in the commodity space at least they are probably overdue for a nice run. I men, they had 25 years of sideways, profitless, horrible markets after the last commodity boom in the 70's.

Presumably the end game for the current run is a similar barren patch for these chaps once supply adjusts to demand (or, failing that, demand adjusts to supply.)

However, per our off-lione discussion, I reckon it could be a bit longer (ie, a year or three) before that happens. And so should we all hope, for if the commodity bull/bubble were to collapse now, than the likelihood of the Japan-style corrosive deflation outcome would increase significantly...and I wouldn't really wish that on anyone- not even Berlioz!

"Cassandra" said...

Minor-league corrosive deflation is a head-and-shoulders better outcome than bank-runs and the whole selling of credit for government security. We should be so lucky to escape with such low-level abrasions.

The commodity thing is itself dependent upon leverage. Some is granted inherently by the exchanges and the nature of the leveraged contracts, but others requires lines and credit. Look around, and I cannot see anyone desperate to lend to finance someone else's speculative pecadillo. So, its not that they should implode, or one wishes them to implode as much as I simply do not see how a piece of gold is better store of value than a a company making toothpaste with a clean balance sheet, good large-scale commercial cotton farm or block of flats in Munich unencumbered, when the former are getting thrashed . Gold may be end up as the last man standing because its the retail hole-in-the-ground of choice , but for connoisseurs of relative value, its already, or nearly past its sell-by date, at investment opportunity opens up in earnest nearly everywhere else.

Anonymous said...

Isn’t this question of the commodity bull breaking just a corollary to whatever Ben and the boys have in store? When they went 50/50 in the beginning, it signaled to me that they were in ‘ depression prevention’ mode, and the radical was to be expected, if one is on the up with Ben’s writings and speeches. Thus, moon-shots for the commod complex. As I meander your logic Cassandra, and agree in the static with it, thoughts of helicopters circle.(Is the market really fully priced for the lengths he seems to be willing to go? Possibly so.) It’s a monetary story, and if Ben tomorrow chooses to go 25bps, I raise my bid on warehouses and residential lots here in the states, and lower my ask on relics and the rest. If he proceeds like he has and seems to intention, then the opposite.

Incognitus said...

There's something weird going on in the NYMEX. ust check the Time&Sales of the last 9 regular sessions. EVERY session had a $1 or more pump into the close.

Also, GS, of "superspike" fame, is now putting out a research report per week with large price targets for oil (150-175-200)...

One also should not forget the Summer'06 and January'07 downward manipulation by GS when they fiddled with crude and gas weighting on the GSCI....

Charles Butler said...

Yeah, OK. So, I had one go at it, then thought the best tactic here would be to go out and renew my ritalin prescription. It helped. A bit.

Are you not failing to distinguish between equity and commodities markets here? And, in the case of the latter, hard and soft? I get the feeling you're looking at wheat (or cotton, really) and railing as a fund manager. Correct me if my wrong, no offense intended.

The specs, in an up or down spiking ag market, have always driven a limit fundamental situation out of control on the basis of normally spurious shit emanating from point-and-figure, crossing of lines, etcetera. Whether that is worse now with math and electronic trading being applied than when it was propelled purely by male body heat might be open to debate.

For farmers, by the way, the consequences of these kinds of markets are more apt to be devastating than otherwise. The 1970's experience was of mortgaging the super-priced family farm to buy the latest Massey-Ferguson and the end result (aside from the failure of that esteemed brand) was a generation-long depopulation of Saskatchewan, reversed finally in 2006.

If it's any consolation, the $250,000 dollars a seat on the Minneapolis Grain Exchange recently changed hands for might indicate the approach of some boundary condition.

Have you tried submitting to 'Tiger Beat'? They use alot of freelancers.


"Cassandra" said...

Charles, this thought experiment was a result not of distaste for the V. Niederhoffers & Covels of the world, but as an attempt to understand why - stylistically - things are the way they are. I believe there is little distinguishment between markets, instruments, etc. Contracts are contracts, prices are prices in the land of the systematic, all are equally playable in this game. My point of departure IS equity because that's my line and that's where I observe and measure these things best, but CTAs/systematic traders are seemingly as happy in equity ETFs, indices and individual names as they are in bellies, cotton or swissie.

Indeed, passion and fury have always been present in the pits. The purpose of this is to provoke discussion and have folks seriously think about what might result from the heady mix of the two under conditions of an altered balance of participant make-up. I believe suuch an altered balance (from historical experience) has evolved in stocks as stabilizers and liquidity providers (market makers, block desks, specialists etc.) have stepped away from making prices in favor of front-running orders, at the same time as concentration has increased relative to floats (or open interest). Noisey markets and prices allow liquidity providers to get back their position (and spread) where trending markets where information is transmitted and "latched-onto" by feedback traders often don't alllow the liquidity provider to see the light of day. The feedback loop here reinforces the withdrawal of liquidity and reinforcement of price trends - whether spurious or not. I think its real, bizarre, temporary (for markets are ultimately adaptive) and probably results in economic loss via reduuced efficiency and erroneous signals that if nothing else, are noteworthy.

Anonymous said...


I agree wholeheartedly with regards to equity markets. And to that you can add the unknown but imaginable changes that ETF's and index funds have added to the mix. But I resist the notion, except in the most general terms, that 'prices are prices'. Ag prices, in the absence of unusual circumstances, are determined on a very uneven playing field - Cargill, say, on one side and a myriad of small operators on the other - and in a world where the deal must be done within a very limited time frame. Under the circumstances, it's doubtful that prices are set very efficiently at the best of times. If anything, they're probably prone to be low rather than high.

My naive guess is that long specs can only have an outsized effect on the procedure under two circumstances. The first is outright crop shortage from drought, or whatever. Textbook squeeze. The second is a situation in which a lack of oversupply coincides with a moment of pricing power all the way up the chain to the consumer. The fact that this 'works' is entirely due to buyers' margins not being threatened by rising prices thus keeping them in the game. This is the position we are in now, and were in the 70's.

Incognitus said...

And there it is again, a $2 crude pump into the closes on the NYMEX ... 10th time in a row this happens.

Mencius Moldbug said...

While I hold no brief for feedback traders, commodity or otherwise, and in fact am quite pleased to see them getting pasted - irrespective of the damage to my own puny book - I must resume my efforts to educate or at least indoctrinate you on the weird relationship between toothpaste and gold.

When you, or another investment professional like you, buy a toothpaste factory or a share thereof, it is not because you think toothpaste, or toothpaste-making equipment, or a block of flats in Hamburg, is a "store of value." Nor is it because you have any no direct demand for dentifrice equipment or German apartments, any more than you have direct demand for a hunk of gold (tacky IMHO).

Rather, you are buying these goods according to CAPS, because you believe they represent a future stream of payments whose risk-return profile according to your metrics is preferable to that produced by, say, government bonds.

These payments are in dollars, euros, etc. Thus if there is any comparison of "stores of value," you are not comparing factories or flats to gold, you are comparing little pieces of paper to gold. Both, of course, are "overvalued" - ie, their demand is not explained by their utility.

In other words, when you move capital between (say) euros and Hamburg flats, you are reallocating savings within a currency. When you move capital from euros to dollars, or ISK, or gold, or whatever, you are reallocating savings across currencies. Quite a different thing!

Again, gold is a currency because it is overvalued, just like official currencies. The fundamental reason for fleeing to gold is that all the official currencies are diluting at a very high rate - considerably larger than the risk-free interest rate on any official currency. If the dollar was a metal and the Fed had the world's only dollarium mine (De Beers, anyone?) the fundamentals would be exactly the same.

The trouble, as with any popular investment, is that not everyone who is buying it understands these fundamentals and is acting on them. The momentum moves into commodities in general are not driven by fundamentals, and have been hitting goods such as wheat which are not natural currencies and cannot sustain overvaluation.

So you see a false correlation between PMs and other commodities - or for that matter, a false correlation between gold (the inevitable winner in any playoff) and other PMs. How long it will take the momo crowd to figure this out is really anyone's guess.

Anonymous said...

Isn't part of the issue with gold that it isn't a promise? How did shares in toothpaste companies fare in Argentina or Russia after their economies went down the drain?

It looks to me like the world's experiences with fiat currencies demonstrate that one has to jump out at certain times. This may or may not be the time to prepare to jump out of the dollar.