Thursday, April 9, 2009

Deja Vu Quantology

Attached is a chart I will discuss tomorrow... For those in the know I would love to hear your thoughts.



(Sleepless nights for Jim Simmons?)
Sphere: Related Content
Print this post

53 comments:

Anonymous said...

~

Let me guess, the August 1-2 pin gives it away. CMBS meltdown?

Anonymous said...

no, woops ok sorry.

Skippy John Jones said...

I will be curious as to what correlation you will draw from the Hedge Fund Market Neutral Index bottoming again. Cursory examination sees no correlation between this index and the SPX or the VIX. So, now I cam curious indeed.

Anonymous said...

Obviously, this is a graph of an exchange traded hedge fund (market neutral).

EconomicDisconnect said...

If this thing is what I think it is then:
-Neutral occupies a small time frame for the hedgies
-Big bets one way or the other are on the way
-Some will be right, some will be wrong
-We get to bail out the wrong ones?

What did I win?

Gigi said...

So for us beginners, this graph shows that hedge funds are not neutral? Meaning that they are very strongly on one side of the market?

EconomicDisconnect said...

If anyone likes puzzles, I had some DNA translation puzzles up on my site for deciphering messages. If you like that kind of thing try:

http://tinyurl.com/dhj9vw

No peeking at the solution posted the next day!

Unknown said...

Hedge funds can't judge relative performance better than average? Alternatively, proof that over the long run, hedge funds in aggregate don't deliver alpha, just beta.

Tyler Durden said...

as an aside, does anyone have a TREPP account I could borrow for a few hours

James said...

Do you think there is a quant dislocation going on under the surface? Possible.

Anonymous said...

I have no idea what this means, but it scares the hell out of me.

EconomicDisconnect said...

Tyler do not go all nigerian scam on us! Whats a TREPP account anyway! You are speaking a different language tonight.

Anonymous said...

i have absolutely no idea what that chart means. but it looks like it bottomed in early 8/07 and mid January 08, which is right before we started massive plunges. so if we are bottoming does that mean we are about to start another big plunge? i actually kind of hope so, i'm getting sick of the idiots on CNBC being so happy

Anonymous said...

the spike up from march 9 till eom played havoc with their mo models.

TPC said...

Looks like the boys are all on the same side of the trade. Will we see a rush for the exits? Will it be a market driver?

Let's have your answer Tyler....

Anonymous said...

Beta neutral does not equal market neutral. Takes unusual recent activity of last month to knock a few ostenbibly market neutral folks around. As an aside, I hope I'm not the only person on this site who is utterly confounded by the last five weeks in the equity markets.

Anonymous said...

"Obviously, this is a graph of an exchange traded hedge fund (market neutral)."

Smartie Pants! :(~

Anonymous said...

Have no idea what the chart means or what a TREPP account is. But I do have to say that they had a freakin' love fest on Kudlow Report today. Is that guy for real? I watch him for the chuckles and to exercise my faux six pack. Does this say anything about the rally today?

http://online.wsj.com/mdc/public/page/2_3022-mflppg-moneyflow.html

Anonymous said...

CMBS yo!!

http://www.trepp.com

Anonymous said...

George Costanza, Hedge Fund Manager.

Paul said...

Im sure a lot of us know how they feel, at least this last time around (I do). They also got hammered by the Nov 20th rally. Heads have been spinning since then. Anyone short liquidity and long common sense knows how they feel. There is plenty correlation here - they are getting killed on the whipsaws. On similar news, make sure you have your REIT, CVX, & BA calls on for Monday, those puppies are going to skyrocket --- and why shouldn't they? I am, of course, kidding.

sweet! said...

trepp, bloomberg, realpoint...let me know.

Anonymous said...

Wake up!

If Market Neutral HFs are losing money since mid-Oct, it means 3 things: 1) they were not MN; 2)they were net long until March 9th; 3) they are net short since March 9th.

Implications -- they will have to cover losses and/or face redemptions, OR, based on the support drawn in this chart, they are as wrong as they have ever been during the past year and about to become 'right' again; i.e. market close to exhaustion.

I want a free one year subscription.

Anonymous said...

My guess about the implications of this chart would be that hedge funds are as short now as they have been in the last 2 years and getting crushed. Shorting the high beta names and buying the index.

Anonymous said...

Tyler, can't wait to hear your thoughts on this tomorrow. Commenters are a bunch of retards tonight and not providing much clarity.

Anonymous said...

mean reversion at its best. Market neutral strategies should provide a return that is comparable to the risk free rate plus a risk premium so the way I look at this is that as the strategies go down in return, risk appetite is coming back to the market.

I wouldn't worry too much about Jim Simmons loosing sleep. He only made himself $4.8b in the last two years and their high frequency strategies are probably kicking ass this year.

Anonymous said...

Tend to agree with Anon @11:06 - this is definitely the sense we are getting in CDS land - furthermore, Anon @11:08 brings up a point we are seeing more and more in CDS markets (i know the chart is for equity market neutral) and that is investors trying to cover short single-name bets with long index hedges in an effort to reduce vol (this is a sucker's bet as when the breakouts occur - the indices suffer from technical pressures and the single-name deltas never 'work' how you'd hoped).
My sense is that the chart shows that MN funds are hurting, liquidity will start to get sucked out of the market, there will be no long/short players to soak up the instutional selling that is being handled by retail buying and MN pairs - we see the correction...

Anonymous said...

thank you anon @11:23, so is this 5 week bull market insanity finally going to be fucking tied to the deteriorating economy everyone other than cramer seems to see.

Anonymous said...

hillbillies: rentech is not eq neutral

this shows nothing except what everyone already knows.

David said...

Stands for Hedge Funds. Equity market neutral. What do I win?

John said...

That is interesting to see that market neutral hedge fund strategies are getting whacked right now. I would have some additional questions though, is it active or passively managed? True MN funds operate accross multiple markets in a vast array of currencies and the forex market has been going haywire for about a year now. I tried looking around but couldn't find exactly how the fund is composed.

If it's not painfully obvious already, T, you've got access to more data than most of us.

It may also be such that the funds repositioning/reweighting is continually backfiring as they aren't able to properly adapt to continually revolving bailout programs/interference plays. Or it could be as others have said that MN actually had a bias and was not following the markets objectively as stated.

In hindsight it looks as if the previous two declines they went net long at the exact moments they should have been short (rather neutral). Currently they must be net short in order to be getting hammered so badly. This must mean the market overall is either due to correct very soon in order to restore market neutrality or if you believe the fund is biased, that it would need to reverse position.

Another way to look at it could be that the large decline (j-M) is in indication that the best performing longs are not performing well and the worst performing shorts are doing exceptional. This would be a major gaffe on the part of the Barclays. The modest bounce is an indication of a return to fundamentals (although today sure would have confused me with regards to fundamentals being traded).

Either way I'm interested to see tomorrows post. For those who don't know a TREPP acct grants access to CMBS/CDO trading, pricing, RM in primary and secondary markets.

Anonymous said...

Could the explanation for the chart be as simple as this...The black swan idea...during emotionally extreme market conditions, rational math based trading and risk management systems break down.

A 50+% rally in a day for a financial stock or homebuilder is a pretty big black swan and hard to hedge neutral against.

Anonymous said...

Forget charts. I've developed a fool-proof trading strategy, which I've been designing and refining for the last few weeks (specifically, since about March 10). It's complicated and esoteric, but since inception it has never failed:

Step 1: Buy SPY.

Step 2: Watch S&P rise.

Step 3: Sell SPY.

It's proprietary and I will be marketing it to investors for a fee, but I'm divulging it to ZeroHedge readers with my compliments.

I'd wish you all good luck, but as you can see above, you don't need it.

"Cassandra" said...

THIS IS THE PICTURE OF STRATEGY UNWINDING & DELEVERAGING'S IMPACT UPON PARI-PASSU POSITIONING ACROSS THE PALETTE OF QUANT-EQUITY LONG/SHORT. JUST LOOK AT THE RELATIVE PERFORMANCE A FEW HIGHLY-SHORTED NAMES (ADVS GMCR AZO ORLY).

THE PUKING OF KERVIEL'S PECADILLO CAUSED A BRIEF UNWIND CASCADE RESULTING FROM THE INEXPLICABLE MASSIVE MOVES (SOCGEN'S PUKING) UNTIL THE CAUSES WERE KNOWN AND UNCERTAINTY PASSED. NOTE THE ACCELERATION INTO THE MAR & SEP 08 QUARTER-ENDS (HF REDEMPTION EFFECTS) AND AGAIN INTO THE CURRENT NADIR AT RECENT QUARTER END.

THERE ARE OBVIOUSLY NUMEROUS FLAVOURS OF STRATEGIES CONTAINED HEREIN, BUT ASSUMING THERE IS SOMETHING LIKE A 70/30 ALLOCATION TO REVERSION-VALUE (CALL IT PRICE SENSITIVE) TO MOMENTUM-EARNINGS (NON-PRICE SENSITIVE) ACROSS THE UNIVERSE, THIS IS PERHAPS A REASONABLE REPRESENTATION OF THE PERFORMANCE OF COMPOSITE POSITIONING THAT REFLECTS VALUE AND SMALL CAP BIASES OF MKT-NEUTRAL PORTFOLIOS.

CHECK OUT "BMNIX EQUITY GP W" FOR THE LISTED POSTER-CHILD OF GETTING TAG-TEAMED BY BOTH YOUR LONGS AND SHORTS AND TAILS. AND IF TO ADD INSULT TO INJURY, THEIR ONSHORE INVESTORS HAD TO PAY TAX ON REALIZED GAINS DISTRIBUTIONS WHILE GETTING SHELLACKED!

Todd M said...

Having run a quant driven paired book, I can tell you that the beatings when they come are relentless. The strategy suffers from complete and total saturation. Go look at the HDS and you will see all the same guys: GS Alpha, LSV, Dimensional, Barclays etc. To monitor them it helps to construct a small theoretical portfolio of stocks. Use a simple regression based on EV to Sales, Growth, and operating margin. The longs should show the above firms as holders. The shorts should be classic "glam" names. My two favorites to watch as a simple pair are NAFC on the long and ELON on the short. NAFC seems to be a much better indicator.

Looking back at my initial failure in Aug of 2007, my first mistake was lack of originality my second was a failure to account for the optionality of equities in certain situations that is hardly captured by traditional valuation metrics. Furthermore, no quant that I know of has effectively dealt with SPEs, SIVs, pension shenanigans and other off balance sheet rigmarole in an effective manner. But this does not mean that long short or market neutral strategies are failures. Comments like that are just too broad and lack clarity to be useful.

Quite simply, imagine a table with ten players, where four of the players control the majority of the capital. There are two time horizons in this game one short term and one long term. The short term one is nearly random with the large players as price setters who employ similar models... The long term one is controlled by lets say the economics of the individual firms over time and is holder neutral. Take this simple game and apply it to our current situation where the large players can and do fail and you have a gedanken experiment for the failure of some of the long short and market neutral quantitative strategies. Just a thought.

Anonymous said...

Todd M. or Tyler,

Can you succinctly explain that in plain English in four sentences or less, for those of us who have not run a quant book.

Many thanks.

Anonymous said...

The last time this HFRXEMN extreme was reached the S&P500 was at 1500+. Since that time the index has declined 700+ points. However, there was a confirmation before the fall. Are we looking for a confirmation? On the other hand, it could also simply mean that there is a big long trend turn coming, one way or the other. The data is too limited to draw a conclusion....

Moderator said...

Why should we be surprised by an inverse correlation of HFR to equities? The bears are going to be wrong for a lot longer than would be suggested.

Anonymous said...

When Quant funds melted down in August 2007, Divid Viniar (CFO of Goldman) called it a "25 standard deviation event." Statisticians were quick to point out that a 25 standard deviation event happens ever on average once every 1×10^185(that is a 1 with 185 zeros after it) days - longer than the universe has existed. Viniar explanation was a fancy way of saying they were wrong.

At the time Goldman Alpha fund was the largest Quant fund in the unknown (and unknown universe). Goldman has subsequently shut it down.

MIT Technology Review had a long essay about what happened in August 2007.

https://www.technologyreview.com/article/19530/

From MIT:

According to most of those to whom I spoke, something like the following occurred this summer. Quants had, in the ordinary nature of their jobs, “shorted” many stocks. Shorting is an arrangement whereby an investor borrows a stock from a broker, guaranteeing the loan with collateral assets placed in what is called a margin account. The investor straightaway sells the borrowed stock; if the stock then declines in value, the investor buys it back and pockets the difference in price when he returns the stock to the broker.

But if the stock unexpectedly increases in value, even for a little while, the investor must either place additional collateral in the margin account to cover the difference or buy back the shorted stock and return it to the broker. CDOs had functioned as the collateral on the quants’ short positions.

When the subprime crunch squeezed the financial markets, the value of those CDOs declined, forcing quants to increase the collateral in margin accounts, buy back the shorted stocks, or both. But in either case, in order to supplement their shrinking collateral, quant funds were forced to sell strong blue-chip stocks, whose prices consequently fell. At the same time, as quants bought back shorted stocks, the prices of those stocks increased, demanding the posting of yet more collateral to margin accounts at the very time that the value of CDOs was suffering.

That the quants were, apparently, long on the same strong stocks and short on the same weak stocks was a result of a number of strategies, pairs trading among them. Another related explanation for the August downturn was that the quants’ models simply ceased to reflect ­reality as market conditions abruptly changed. After all, a trading algorithm is only as good as its model.

Unfortunately for quants, the life span of an algorithm is getting shorter. Before he was at RiskMetrics, Gregg Berman created commodity­-trading systems at the Mint Investment Management Group. In the mid-1990s, he says, a good algorithm might trade successfully for three or four years. But the half-life of an algorithm’s viability, he says, has been coming down, as more quants join the markets, as computers get faster and able to crunch more data, and as more data becomes available. Berman thinks two or three months might be the limit now, and he expects it to drop.

---

So it seems that another 25 standard deviation event is now unfolding. And yes, Simons (not Simmons) will always be at the head of the list when Market Nerutral Funds are going awry (like your chart shows). He now has the place that Goldman's Alpha had, he is the largest quant fund in the universe.

The question is why? Here is my (pure) guess.

Are the Quant funds long the credit of financials and short the equity of financials?

Many have assumed that when the recovery comes, credit goes first and the stock market follows (this has actually been the case for several years now.) The S&P 500 Financial Index is up 40% since March 9 and the credit have barely moved.

Unknown said...

"Are the Quant funds long the credit of financials and short the equity of financials? "

Like what happened with GM years ago but on a larger scale?

Anonymous said...

I run a small quant equity hedge fund, and these past 4 weeks have been pretty ugly indeed. But I think you can simplify the explanation: low-quality stocks that were widely shorted have gone on a moon shot during this mother of all bear market rallies. If you want examples just look at HOG and ETH... common shorts, awful fundamentals, crippling debt, and rich valuations. I suspect Tyler is going to make the point that this "negative alpha" from common-sense fundamental factors suggests this rally is completely fake, and I would agree with him. Unfortunately, if you are a manager who relies on an active-alpha short book, things are getting very tough, and the government's commitment to not allow firms to fail and to monetize debt has made it quite dangerous to short-- and the market flows are reflective of that new reality.

Anonymous said...

"Are the Quant funds long the credit of financials and short the equity of financials? "

Yes, by boatloads.

DumbAndDumber said...

The stock market is turning out to be the Best of the Worst, in terms of places to park worldwide cash.

Market goes up.

Quants/hedgies cover shorts using the index, market gets pushed up by the index.

Everything looks rosie to retail. Retail starts diving in, market goes up.

Quants/hedgies cover shorts using the index, market goes up pushed by index.

Everything looks even rosier to retail. Retail buys even more, market goes up.

The Law of Unintended Consequences.

Yikes! What's Barney Rubble going to do about this? All algorithm theorists and quant coders will be sent to Guantanamo.

Anonymous said...

The chart says short stocks. Simple.

Anonymous said...

Like what happened with GM years ago but on a larger scale?"

No, that was 2005 when F and GM were downgraded and the CDS correlation trade blew up.

This is more like August 2007 when Goldman's Alpha fund was coming apart.

Anonymous said...

Like what happened with GM years ago but on a larger scale?"

No, that was 2005 when F and GM were downgraded and the CDS correlation trade blew up.

This is more like August 2007 when Goldman's Alpha fund was coming apart.

Anonymous said...

I run a small quant fund as well and this has been one of the best years we have had in a long time. As everyone moves into the high frequency and relative value world, we have moved away into lower frequency models based solely on the mathematics underlying the time series (in this case price). The problem is that if you do what everyone else does the profits go away. An MIT degree, an article about you in Automated Trader and an in-house built Cray supercomputer architecture does not necessarily make a good trader or fund.

Simpler CTA type systems that act much like a trader would will always be successful in the long run because there will always be behavioral arbitrage opportunities as long as discretionary traders and the likes of Jim Cramer are around. The automation also keeps you from harming you.

We have been using the same systems for many years with very little variation. Our search algorithms have gotten faster and more efficient but 10 i7 machines with distributive processing capabilities are much more than we need. The lower frequencies also save the need for any latency critical infrastructure. However, we don't run billions of dollars and are not hindered by size so for us simple is better.

Anonymous said...

anon 2:15 are you hiring?

Anonymous said...

The scale on this graph makes it look worse than it is. This index is down about 3.7% YTD, and down about 8.6% from peak (7/07). Compare that to other asset classes. Recent underperformance is probably driven by a quality bet the factor-based managers added after it did so well last year. Deleveraging combined with commonly used risk models (BARRA) are also playing a role.

DumbAndDumber said...

anon @ 2:48.

While you're waiting why not develop a few algorithms that have proven results. Depending on your backtesting results you may be able to shop them around to get yourself bootstrapped.


http://code.google.com/p/hudson/

Anonymous said...

I don't understand why JS would lose sleep?

His closed Medallion fund makes a mint either way the market moves.

Given his historical returns from the Medallion fund, I'm not sure why he bothers running his institutional funds.

Mallory said...

Is there a way to follow this without a Bloomberg terminal? The index is very interesting.

Unknown said...

http://www.google.com/finance?q=NASDAQ%3AHSKAX