Quant Funds & The August Market Turmoil
John Mauldin's Outside the Box

Blog Subscription Form

  • Email Notifications
    Go

Archives

What really happened last August? There was blood in the street for many hedge funds, while others did ok. But in this week's Outside the Box, Jon Sundt from Altegris Investments (and my US partner) dives you the behind the scenes details of what was going on inside the trading rooms of various quantitative hedge funds. It makes for interesting, if not sobering, reading.

I think you will find this analysis helpful in your own efforts to analyze your investment managers. How much real risk are your managers taking? Can you tell just by the performance numbers? Jon suggest different ways to look at the risk in your portfolio.

Jon is the president of Altegris Investments. He has been researching and analyzing hedge funds for many years, along with his partners and an extensive research team at Altegris. I am proud to partner with his firm, and happy to present you his essay ad this week's Outside the Box.

John Mauldin, Editor
Outside the Box


ADVERTISEMENT
Oil over $100 a barrel.
Today's energy market is hard to maneuver... it takes an expert to find the right companies to invest in. Oil and gas, uranium, geothermal energy are all sectors where you can make a fortune or lose your shirt. The Casey Energy Speculator sifts through the stock market for you and separates the wheat from the chaff. Give Casey Energy Speculator a no-risk try with our 3-month money-back guarantee. You'll be glad you did!



Quant Funds & The August Market Turmoil

In my last newsletter, I wrote about the relatively obscure group of hedge funds that speculate in subprimes and how, for the most part, the average hedge fund investor avoided them.

It is interesting to note that the issue wasn't the hedge funds that invested in subprimes. Rather, the issue was subprimes themselves, and the now infamous lending, packaging, re-packaging and selling of subprimes that followed.

Shortly after subprimes made headlines in July/August, as defaults shot upwards, another group of hedge funds made the news. This time, it was the quant funds. From July through August, some of these quant funds lost billions.

What is a Quant Fund?

Like stock fund managers, there are various breeds of hedge fund managers. Among equities, fund investments may be concentrated in companies or indexes of a particular growth stage, industry, or geographic area, such as Growth vs. Value, Tech vs. Consumer stocks, International vs. Domestic, etc. Many hedge funds also specialize in a particular field and are typically categorized as Global/Macro, Long/Short, Distressed, Quantitative, Market Neutral, etc.

Licensed for use from the New Yorker Magazine.

In the hedge fund world, the label "quant fund" has a distinct meaning, quite different than your plain vanilla long/short fund. But unless you knew better than most, differences may have been hard to distinguish until this fateful August.

The funds that got the most press were market neutral quant funds characterized by the "statistical arbitrage" or "algorithmic trading" models they used. These models allow a computer to scour historical price data for relative value inefficiencies between stocks, futures, currencies, or fixed income securities.

To better explain what happened to quant funds, I will tell a fictional story that is roughly based on the facts as we understand them at Altegris. I will change the names to protect the innocent (and avoid any litigation!).

Let's create a hypothetical quant fund called the PhD Fund. The fund is owned by a big name Wall Street firm and is marketed on the street to wealthy individuals and institutions under the banner of a "market neutral" fund. This means that the net exposure for the fund is zero, or, in other words, the dollar amount of long positions in the portfolio is offset by the dollar amount of short positions.

The PhD Fund employs dozens of "propeller heads" (a hedge fund moniker for our mathematically inclined friends in quant shops). They build computer-based models that try to find and trade overvalued and undervalued stocks. Some of these models are longer term but some are short term, and so in order to trade in and out of these markets, these funds need to trade large, liquid positions.

The PhD Fund chooses to play in the US stock market because of this market's breadth and depth of securities. Let's assume that PhD Fund has dozens of measures for stocks in certain sectors and their models are constantly being built and augmented. They have income models that look at EBITDA, quarterly earnings, growth and more. They have balance sheet models that look at debt to equity ratios and book value, among other figures. They have technical models that look at short term momentum, daily volume, open interest and daily tic-by-tic trade data. You get the picture. These funds employ highly paid Gepettos, pulling the strings on computer models and trying to create money out of historical data.

Quant Clusters

The PhD Fund's goal is to find as many market inefficiencies as possible. To do this, they might look at constructing a large group of stocks, called "clusters," to analyze. The secret sauce is how the computer models construct a cluster, perhaps looking at daily price, balance sheets, analyst estimates, short interest, momentum, etc. Once these clusters are defined, the computer model will act by selling the overvalued and buying the undervalued stocks within a particular cluster. Like many quant funds, the PhD Fund uses an electronic trading platform, and holds close to 1000 securities long and short at any given time.

Here is a very basic example of how one of its systems might work: The PhD Fund culled historical data on the S&P 500 for the past five years. The model observed that every company with a market capitalization of $500 million to $600 million has fairly stable returns; the most a price will change in one day is around 1%. The model also showed that when any of these stock prices become consistently more volatile, they generally move in the downward direction. Let's assume there are 60 stocks that meet this $500-$600 million market cap criterion. Within this group, 20 of the 60 stocks showed significant daily volatility over the last week. One day some stocks are up 5%, the next day some are down 2%. Since the model's data shows that increased stock price volatility leads to price declines, PhD fund will likely short the 20 stocks and stay long the remaining 40.

What I just described was one example of clustering based on past historical relationships that are believed to repeat. Because the PhD Fund is a savvy shop, they have 10 different "cluster" systems in their model. What really excites them is that each system is unrelated to the other. Some have different time horizons, mean reversion, short term momentum, sales growth, free cash flow, etc. PhD therefore assumes that the Fund has low correlation among all their models. Makes sense, right?

This is a very simplified explanation but it will suffice for the context of this letter. I have left out optimization, normalization of factors, liquidity constraints, risk monitoring and automated execution among others. In reality, these quant funds can be far more complex but the above explanation paints a generally representative picture.

Hazard of Leverage and Size

Our PhD Fund has run its models over the last five years and made a killing. The propeller heads are rich. They have also found that because of the low volatility in the market the last five years and the low correlation within their market neutral system, they can leverage their Fund. This low volatility lulled many quant shops into higher and higher leverage. The lack of any recent blowups or spikes in volatility made them feel immune to market jolts.

The models were formerly picking up quarters, but now they are picking up nickels. Because the models would need to pick up more nickels to make the same amount of money, many turned to leverage for help. The PhD Fund decided to lever 8:1. For every $1 million the Fund put forward, it borrowed enough to have $4 million for its long book and $4 million for its short book, all the while keeping its "market neutral" label. Its net exposure was zero ($4 million long plus $4 million short), but its gross exposure is 8x. It was genius. The PhD Fund amplified returns, all the while keeping its market neutral hat on. It had $1 billion under management, before leverage. With leverage, its assets were $8 billion. It's no wonder that as volatility in the markets remained low over the past five years, quant fund assets soared.

ADVERTISEMENT
World Currency Booklet!
With the advent of the Philadelphia Stock Exchange's new cash-settled world currency options and futures, trading in the world of foreign exchange has taken on a new dimension! Learn choice of markets and accounts, contract sizes, specifications, style, and more!

Find Out More...


Plain Vanilla Long/Short Fund

For comparison, let's create a hypothetical new long/short fund that uses a familiar strategy in the hedge fund world. Let's call it the Plain Vanilla Long Short Fund, or simply the Vanilla Fund.

The Vanilla Fund is a hedge fund with around $400 million under management. It has an experienced research team that evaluates fundamental measures of a company's stock (bottom-up research) as well as overall industry trends (top-down research). The team buys what they believe are undervalued stocks and sells what they believe are overvalued stocks. They use computers for some input and risk monitoring, but they also talk to "the street", monitor quarterly calls with stocks they follow and glean insight from a variety of industry contacts and experience, before buying or selling a position. The Vanilla Fund's experienced research team has spent years analyzing stocks. Some might call it fundamental research. They trade 50 positions long and 50 positions short. They explain to you that they will keep their book market neutral, so their net exposure is zero. They do this by using REGULAR MARGIN available for most brokerage accounts. The Vanilla Fund borrows $1 million for every $1 million dollar invested, meaning it uses $1 million to go long and $1 million to go short, for a gross leverage of 2x. With leverage, the Vanilla Fund has $800 million under management.

Time for You to Invest

Let's pretend it's January of this year and you are thinking of investing in a hedge fund. You feel the market may be overvalued, so you want a hedge fund that actually hedges (you'd be surprised how many don't). On your desk are the Vanilla Fund and the PhD Fund. Both are market neutral and go long and short US equities. Both have stellar track records, a low correlation to the S&P 500, and reasonable performance in up and down markets -- after all, they can short and are market neutral.

So you write a check for a million bucks and invest in the PhD Fund, because after all, the head guy went to MIT and there are too many smart people in that Fund for it to screw up.

It is estimated that market neutral quant funds have risen nearly 60% in the last two years alone, to $96 billion as of June 30, according to Hedgefund.net. These figures do not reflect leverage. The astonishing growth reflects both investment gains and new money.

The Vanilla Fund and the PhD Fund both have zero net exposures...for every dollar long they have a dollar short. Combine this with their performance, and they look pretty similar, right? Wrong.

The difference can be found by asking the question: What is the Fund's gross exposure?

Here the differential is huge: 200% for the Vanilla Fund compared to 800% for the PhD Fund. Gross exposure shows just how leveraged these funds are: 2x versus 8x in this case.

Growth vs. Value Statistics*
08/07/2007 to 08/09/2007

Large-Cap Stocks
(Russell 1000® Value Index)

-1.04%

Small-Cap Stocks
(Russell 2000® Growth Index)

+3.24%

The Perfect Storm: August 2007

Now August rolls around. You're invested in the PhD Fund, and out of interest you are still monitoring Vanilla Fund. Unbeknownst to you, the period of time between August 7 and August 10 was the perfect storm for quants. In what some have described as a "standard deviation of 25" event, there was a sharp reversal of the relationship between the stocks the quants were long and the stocks the quants were short. It was a violent four-day event, with higher quality value names dramatically underperforming lower quality growth names as you can see in the graphic to the right. The relationship between the individual stocks was even more severe. For some quant funds, these events nearly put them out of business.

So what happened to cause such a cataclysm for quant funds in August? To recap, the subprime contagion was worse than expected, creating a general uncertainty in the market. Volatility increased as stock prices moved all over the place intraday. Hedge funds across the board began to take risk off their portfolio, reducing their exposure to the market. In August, both sophisticated and novice investors alike took their money out of the market, selling what they could in order to raise cash. As you can imagine, it's easier to sell the good stuff than the bad stuff. The most liquid, value-oriented, reputable stocks were sold by panicked investors because it was easy to do.

Models at the PhD shop can't predict this kind of market panic. With historical relationships between securities, whether based on price, statistics or some other measurement not holding up, the quant models lost their predictive value.

To make matters worse, the PhD Fund lives in the quant world where many papers have been written and many competing propeller heads have opened up similar shops. What the PhD Fund didn't know was there were dozens of competing quant funds all trading similar strategies. Many quant funds identified very similar inefficiencies, so many were buying and selling the same positions. Remember, volatility has been very low for five years and value stocks have outperformed growth. This means that many of these giant quant funds were long value and short growth, just like their models told them to be. And they all ran for the door at the same time. So they had to SELL THEIR LONG VALUE and BUY THEIR SHORT GROWTH. Value got crushed and growth rallied. They got hurt on both sides of the trade. Their hedges (shorts) went against them.

Add the fact that all these quant funds were very large, highly-leveraged, and carrying similar positions, and you have a recipe for disaster. It wasn't just the strange behavior of the market, but the LEVERAGE and SIZE of the funds trying to get out. The PhD Fund is leveraged 8x, compared to the Vanilla Fund, leveraged 2x. The PhD Fund suffers a 4% loss on their longs, and a 4% loss on their shorts before the leverage. Because of the leverage, you have to multiply that by 8, for a 32% loss!!

Our friends at the PhD Fund are in a panic. It's mid August and all the red lights, bells and whistles are sounding off. It's "Defcon 5" and their computers are buying and selling stocks in a flurry as stops get triggered and new signals are generated. At the close of business, August 15th, their Fund is down over 30%. They are in "shock and awe." An event that their computers told them should only happen once every 100 years has happened (funny how often I hear that). A propeller head in the back room says to hang on, because he remembers the Russian Debt Crisis, and how, if people had only hung on, they would have been fine. The CFO says they have to hit the reset button, liquidate everything and reassess. No one knows what to do, and the models are worthless.

ADVERTISEMENT
Win With Options In Any Market
Trade options like a pro for profit and protection. Free Guide.

Find Out More...


Decision Time

Our friends at the PhD Fund decide to hang on, and lo and behold, by the end of August, their fund recovers most of the losses. The reversion to the mean happens in the nick of time, and the stock-price relationships that were out of whack finally fall in line.

The decision to hang on was purely arbitrary, with no predictive model able to give comfort. When September 1st rolls around, the Fund posts a modest loss. The propeller heads at the PhD Fund all look in the mirror and reassure themselves that their genius is intact and the 30% intra-month loss was nothing more than an aberration. They deride their CFO for suggesting they override their system and sell everything.

Quant Fund Losses
August Performance

Funds That Did Not Liquidate

Fund A

0.4%

Fund B

-0.4%

Funds that Liquidated

Fund C

-20%

Fund D

-22.5%

Although the fund names have been withheld, the returns above represent actual performance.

Meanwhile, across town, a competing quant fund called the Braniac Fund feared the market would worsen and hit the reset button, liquidated all positions and went to cash. They lost faith in their models and believed they could easily lose more money. In an effort to get some breathing room and reassess, they overrode their model: human intervention. The Braniac Fund posted a -30% loss for August and were lambasted by their peers for lack of faith in the computers.

And the Vanilla Fund...well, they lost -8% intra-month. After their risk management kicked in, they posted a -6% month for August. No red lights, no panic, no excess leverage, just prudent risk management. They are now easing back into stocks, and welcoming the volatility.

You, as an investor, are mercifully unaware of any of this. You had no idea that mid-August the PhD Fund was down 30% and the propeller heads were levitating and on the verge of implosion. All you know is at month end the PhD Fund posted a -1% loss and you find out the Vanilla Fund posted a -6 % month. You mumble on the way to your foursome that you are glad you didn't invest in Vanilla, and you are reassured when you receive a letter from the PhD Fund that, although they fired their CFO, they hired four new propeller heads.

Best regards,

Jon Sundt

Jon Sundt
President and CEO
Altegris Investments, Inc.
Trusted Alternatives. Intelligent Investing.SM


1 WSJ.com Web Link
2 What Happened To The Quants In August 2007? Amir E. Khandaniy and Andrew W. Loz First Draft: September 20, 2007

* Source: Russell.com


Your thinking more black swan events are in our future analyst,

John F. Mauldin
johnmauldin@investorsinsight.com


Disclaimer

John Mauldin is president of Millennium Wave Advisors, LLC, a registered investment advisor. All material presented herein is believed to be reliable but we cannot attest to its accuracy. Investment recommendations may change and readers are urged to check with their investment counselors before making any investment decisions.

Opinions expressed in these reports may change without prior notice. John Mauldin and/or the staffs at Millennium Wave Advisors, LLC and InvestorsInsight Publishing, Inc. (InvestorsInsight) may or may not have investments in any funds, programs or companies cited above.

PAST RESULTS ARE NOT INDICATIVE OF FUTURE RESULTS. THERE IS RISK OF LOSS AS WELL AS THE OPPORTUNITY FOR GAIN WHEN INVESTING IN MANAGED FUNDS. WHEN CONSIDERING ALTERNATIVE INVESTMENTS, INCLUDING HEDGE FUNDS, YOU SHOULD CONSIDER VARIOUS RISKS INCLUDING THE FACT THAT SOME PRODUCTS: OFTEN ENGAGE IN LEVERAGING AND OTHER SPECULATIVE INVESTMENT PRACTICES THAT MAY INCREASE THE RISK OF INVESTMENT LOSS, CAN BE ILLIQUID, ARE NOT REQUIRED TO PROVIDE PERIODIC PRICING OR VALUATION INFORMATION TO INVESTORS, MAY INVOLVE COMPLEX TAX STRUCTURES AND DELAYS IN DISTRIBUTING IMPORTANT TAX INFORMATION, ARE NOT SUBJECT TO THE SAME REGULATORY REQUIREMENTS AS MUTUAL FUNDS, OFTEN CHARGE HIGH FEES, AND IN MANY CASES THE UNDERLYING INVESTMENTS ARE NOT TRANSPARENT AND ARE KNOWN ONLY TO THE INVESTMENT MANAGER.

Communications from InvestorsInsight are intended solely for informational purposes. Statements made by various authors, advertisers, sponsors and other contributors do not necessarily reflect the opinions of InvestorsInsight, and should not be construed as an endorsement by InvestorsInsight, either expressed or implied. InvestorsInsight is not responsible for typographic errors or other inaccuracies in the content. We believe the information contained herein to be accurate and reliable. However, errors may occasionally occur. Therefore, all information and materials are provided "AS IS" without any warranty of any kind. Past results are not indicative of future results.




Posted 11-19-2007 3:18 PM by John Mauldin
Related Articles and Posts