Is Quantitative Trading alive or dead?

Scores of quantitative hedge funds have taken hits over past year that make NFL hits in Sunday football games appear to be minor league. Many have questioned if there is a future to the quantitative approach to trading the market, all the current models appear to be completely discredited. Yet to entire industry continues to drive forward, new algorithms are being created, however most are still focused on wringing raw profit out of the market based on differences in data input.

Similar to the movie Groundhog Day, repeating the same mistakes will only lead to a cycle of the same results. The continuing flawed premise is the over-optimization of models to focus on profit while ignoring risk.

A recent study by the CFA Institute outlines many of the issues with quantitative models used by Hedge Funds. Many factors have led to a complete decay in performance, leaving funds scrambling to unwind exposed positions. The information reveals that profits opportunities for quants are harder to find and exploit, while the updated strategies expose firms to increasing risk. The concept that firms will trade strategies that increase risk while possible gains are minimized bodes poorly for the industry.

A couple of HingeFire briefs have touched on this subject in the past (see Is there a future for Quant Funds and Quants search for new math). The complete 100+ page study from the CFA Institute is available for free in PDF format and is very educational. It can be downloaded here - http://www.cfapubs.org/doi/pdf/10.2470/rf.v2008.n2

Is there a future for Quant Funds

It appears the Quantitative Funds have been stranded on the lee shore, and a crowd is desperately trying to find a way to shove the ship off the beach. Even if re-floated the quantitative industry is mired in a cross-wind about its future direction; not knowing whether to tack or jibe in order to catch air in the sails.

The recent draw-downs during the credit crunch have raised the question if the mathematical wizards can navigate the waters; or if the conjurers would be better off safely ashore teaching theory at a university. Is there even a future afloat for these funds? The trends of lower cost IT, more powerful computers, mountains of instant data, and new modeling techniques can play both for and against these funds. The bottom line is that a proper definition of risk still haunts the players.

Earlier HingeFire articles (see Quants search for new math) discussed the confusion as these funds search for a new formula that will pull alpha out of the market while not leaving the funds sunk every few years. The current methods have proven to be more akin to gambling than investing.

Advanced Trading recently addressed the fund dilemma - Quants Searching for Alpha: Do the Pros Outweigh the Cons?

Quants search for new math

For many years, quantitative hedge funds were hailed across the Wall Street community as inspired leaders in the next generation of finance. All the old rules about valuation, risk control, and proper evaluation were all relics of the past. The next phase of the “brave new world” of finance was being defined by these mathematical geniuses, hailed by their colleagues and the press as brilliant.

Maybe the Long Term Capital Management failure in 1998 should have served as a warning about the train wreck that was ahead. (BTW John Meriwether, the most well known of the “Geniuses” who lost billions with LTCM, has just blown up his new hedge fund). Perhaps the notion that all these quantitative hedge funds were pursuing mirror strategies with no risk control should have waved a warning flag.

Over the past six months, Hedge Funds have been closing shop at the rate of over three per week. Among the most severely impacted are the funds pursuing quantitative strategies. Many are victims of the extreme leverage used to trade the risky computational strategies.

When discussing these distressing fund blow-ups the press has started to use terms like “bet on the market” instead of focusing on quantitative math when discussing the latest victims. Most quantitative funds are suddenly viewed as pouring money down a rat hole, with the perception that their “math” was doomed to blow-up suddenly regarded as the foreseeable outcome.

Earlier Hingefire articles (see The Redefinition of Risk and Quants Meet Reality) outlined many of the problems with quantitative strategies deployed by the Wall Street wizards. It was simply a matter of time before a multi-sigma event took out many of the funds. Black swans do exist, and strategies that do not deploy reasonable risk control measures all meet the inevitable brick wall.

So what are the quants doing now their house of cards has come crumbling down? The mathematical wizards are searching for new models to replace the old ones that have gone up in smoke. Maybe this time they should throw in several cups of risk control into the recipe.

A recent article in Alpha Magazine outlines The New Math. Hedge funds are pressing ahead with new mathematical concepts for trading. The math wizards are hunting for new arcane magic that will give them an edge on the street; molecular physics, mathematical linguistics, artificial intelligence, behavioral response, and any other possible concept are on the table. The funds are also looking beyond traditional financial instruments into other asset classes. The bottom line is that most quantitative strategies are still searching for instruments which are mis-priced, yet risk control is still not a priority.

While the community has to give credit to the quants for pressing the boundaries and searching for new math that will provide them a computational edge on the market; maybe it is simply time for the wizards to focus the effort on risk control in their models to ensure long term market trading success. Even at the cost of several basis points of profit in their models and a reduction of leverage; having a strategy that does not blow up the firm every few years must have some inherent value. At minimum, it will eliminate the need to update their resumes regularly and search for a new fund to hang their math diplomas at.

Black-Scholes: Is it simply wrong?

The Black-Scholes pricing model has already been battered by corporations that state the model does not properly reflect the pricing of employee stock options. The gripes of many technology companies include the realities that these employee options are not liquid, have long vesting periods, are lost when an employee leaves the firm, and other factors making the value derived by Black-Scholes absolutely inaccurate. This naturally leads to further imprecision and financial engineering in corporate filings – which does not benefit shareholders.

Now Black-Scholes is under attack from another front – the mainstream financial community. Most traders have ignored the Black-Scholes model for years; finding it next to useless in a fast-paced financial marketplace. However the model is still ingrained in the minds of pension administrators, hedge funds manager, and other institutional executives. There is growing evidence that the model is simply wrong, the most significant problem being that Black-Scholes was never designed for any type of extreme market situation.

Is it time to retract the Nobel Prize awarded to Myron Scholes and Robert Merton for their work in creating the model? Nassim Nicholas Taleb, the author of The Black Swan and Fooled by Randomness appears to believe so. Most of Wall Street is not ready to take this extreme leap however. Many are simply hoping for an updated model that can properly reflect risk.

The current Black-Scholes model views that a multi-sigma event will only occur once every million years, in reality market extremes are much more common. With events that can easily wipe out most risk models occurring regularly within eighteen year periods it is clear that Black-Scholes and other models are broken.

The recent sub-prime crisis should serve as an example to the financial community why proper modeling is needed; rather than greed-driven models focused on putting the largest profits into the pockets of Wall Street firms in the shortest possible time period. While a homeowner who lost their home may not understand the modeling of a credit crisis, Wall Street financial engineers should have an understanding of the obvious risks in their products before billions of dollars are taken as losses.

The panic of October, 1987 should have called the relevance of Black-Scholes into question. A model does not work when investors attempt to sell and nobody will buy. When liquidity is lost, most currently utilized models are worthless. Still the Wall Street community clung to the Black-Scholes model.

Most alarming, if Black-Scholes does not properly model options then trillions of dollars' worth of securities have been priced over the past years without regard to the possibility of multi-sigma events in the markets.

More press is regularly appearing in mainstream financial journals questioning the Black-Scholes model; the detractors will no longer be mute. This trickle is likely to become a roar as Wall Street firms continue to take losses over the coming year which reflects their inability to correctly model risk. At some point, Black-Scholes will be relegated to the financial bit-bucket of the modeling past… only leaving the question if will it occur soon enough to avoid additional pain.

Inside Wall Street’s Black Hole

Will 130/30 Funds hold water?

As pointed out in a recent article in Investment Dealers' Digest, 130/30 Strategies are Set to Gain Traction. Hedge funds and major brokerages have been heavily marketing these funds to institutions such as pension funds. The recent tumble in stocks has only increased the marketing blitz and associated claims that these strategies will squeeze out excess alpha in both down and up markets.

As outlined earlier, 130/30 funds allow managers to short-sell up to 30% of their portfolios, and use the proceeds to buy an extra 30% long. The funds both use leverage and short-selling. The current market size is estimated to be $50 billion, many analysts expect the funds in 130/30 products to grow to $1 trillion over the next few years. However it is an open question if these funds actually can deliver on their promises, or are simply a scheme to increase the fees generated for brokerages and hedge funds.

The usual sales pitch involves presenting quantitative back-tested models and presenting results to investors which demonstrate the increased alpha. Naturally these models have the advantage that the managers can easily tweak the selection criteria to provide the results desired over the time frame. The sub-prime CDO fiasco should provide ample evidence that “quant” models that work in back-testing can easily blow up when applied to a real market. The 130/30 models used by fund managers face a similar risk of immediate under-performance.

The stock market performance in January was dismal, NASDAQ was down by over 9.9% and other indexes also suffered significant declines. Most people would assume that the performance of 130/30 funds would shine in this type of environment, and many would be near the top of performance lists. The reality is that the performance of the 130/30 funds in January was effectively lackluster in squeezing excess alpha out of the market. There is no data that demonstrates any type of significant advantage in the market when all the factors are taken into consideration. In fact, the results from publicly available mutual funds with 130/30 strategies show that they have underperformed the indexes as outlined in Verdict Still Out on 130/30 Leveraged Funds at TheStreet.com (a performance table is provided here). This can hardly be considered “squeezing excess alpha out of the market”.

Most institutions would have been better off allocating a portion of their money to short funds while placing the majority of their funds (80%+) in long opportunities; if they wanted to achieve better performance with lower fees. Veryan Allen touches on some of these issues in his 130/30 overview. Most investors would be better sticking to traditional products than using 130/30 funds according to Morningstar.

A number of industry specialists would state that 130/30 funds simply limits the long side returns while holding the fund steady in down markets. Another significant issue is that the fees charged for these vehicles are often over-sized compared to other types of funds that institutions can use to achieve comparable exposure. Obviously, a race is on by large financial services firms to acquire more institutional assets in this down market. The 130/30 funds serve as a compelling story, and also deliver out-sized fees to the large financial institutions. One industry quip stated that the collection of fees for a 130/30 fund are usually 0.53% greater than the combination of simply buying equivalent long and short funds. Pension & Investments Online states that a pension fund typically pays 50 to 75 basis points for an active U.S. large-cap strategy, it pays about 25 basis points more for a 130/30 strategy. PIonline views that 130/30 strategy funds are simply a payday for money managers.

Another industry issue is that long managers simply don’t have shorting experience. Some funds will attempt to bring in money managers on board with this type of experience while others will just pick stocks they view as weak as their short candidates. Simply picking the bottom 10% of screens means that many times the manager is selecting the 10% anticipating a rebound; due to their lack of understanding of the criteria that should define a short candidate near the price peaks rather than troughs. For example, the majority of basic fund screening and ranking strategies would have money managers shorting banks now (when they are likely near their bottom) rather than mid-2007 (when they desperately deserved to be shorted). As outlined in the IDD article, "Shorting is a rare talent to begin with, and it's hard to consistently profit on the short side," says Chris Wolf, managing partner at San Francisco-based funds-of-funds Cogo Wolf.’ It is doubtful that very many of these funds will exceed the 130/30 index defined by Andrew Lo at MIT.

An additional concern is “negative carry”, this is the spread in the borrowing cost required to put on the combination of long and short positions. This cost of leverage immediately creates a negative alpha that the fund must overcome to even arrive at par performance. Thomas Kirchner, the manager of the Pennsylvania Avenue Event-Driven Fund (PAEDX), provides an excellent explanation of this issue in his Negative Alpha is Built Into 130/30 Funds commentary.

In some of the cases, the interest on the combined long-short position goes directly back to the sponsoring institution; the large financial services firms are apt to view this as simply another revenue stream and use it as a mechanism to squeeze extra money out of investors. Nor are the disclosure documents very clear on the size of the payments. This entire situation is definitely self-serving and effectively acts as yet another fee targeting customers.

The combination of high fees, minimal short-selling managerial experience, and interest costs are likely to doom 130/30 funds to be underperforming entities. Certainly the marketing blitz from large financial services entities will drive these funds into institutional holdings such as pensions. However individual investor should avoid these funds, and seek other alternatives to create a properly diversified portfolio that can deliver respectable returns over time in both up and down market conditions.

The Redefinition of Risk

Summer 2005 – Wall Street
“All swans are white” shouted the leader of the investment banking crowd.
“and what if a black one shows up,” countered the risk manager.
“Do you know how much money we are making on these CDOs? How can there be any risk. Risk is a thing of the past. This is the new generation of banking, risk is distributed," cried the room in chorus.
“The black swam is stalking you, and it is just a matter of time till it makes its appearance,” shouted the risk manager over the din.
“The quants have modeled this stuff over the last ten years. They assure me there is no risk and we are going to make a ton of money in fees. Don’t rock our rice bowl,” stated the lead derivative banker as the crowd rose to leave the room, “Let’s go make some bonus money!”
The risk manager sadly shook his head while staring directly at the black swan that was obviously sitting on the table.

The risk manager resigned the following week and was quickly replaced by one more pliable to the demands of the derivative structure crowd. The figurine of the three monkeys on the new risk manager’s desk was an evident sign of the impending future.

The combination of greed, poor modeling, and deliberate disregard of proper diligence practices has come home to roost in one of the largest catastrophes ever experienced by Wall Street. The events of the last six months are an outline of a history lesson that will be taught in business school classrooms a hundred years from today. The trail of wreckage is widespread and unprecedented. Adding liquidity from the Fed was not able to unwind the credit crisis, and the industry was not able to bail itself out. Many major financial institutions had to turn to foreign sovereign funds to provide cash to avoid being effectively left insolvent from a capital ratio perspective.

Value at risk is dead, a new model is needed that properly accounts for derivative risk on Wall Street. The new representation needs to take into account the expectation of six-sigma events and disregard the pressure from bankers focused on greed over common sense. The concept that firms can solely account for risk by defining the dollar amount at risk on any given day with 95% confidence is a failed relic of the past. The quantitative models from Wall Street wizards that cherry-pick ten year nonvolatile timeframes as evidence that no significant losses can ever occur for derivatives has been debunked. The fantasy in the brave new world of banking that risk can be distributed in the derivatives market and nobody will ever be left holding the bag has obviously crumbled into dust.

A recent article (Death of VaR Evoked as Risk-Taking Vim Meets Taleb's Black Swan) outlines these concerns with this opening:

"The risk-taking model that emboldened Wall Street to trade with impunity is broken and everyone from Merrill Lynch & Co. Chief Executive Officer John Thain to Morgan Stanley Chief Financial Officer Colm Kelleher is coming to the realization that no algorithm or triple-A rating can substitute for old-fashioned due diligence.

Value at risk, the measure banks use to calculate the maximum their trades can lose each day, failed to detect the scope of the U.S. subprime mortgage market's collapse as it triggered more than $130 billion of losses since June for the biggest securities firms led by Citigroup Inc., Merrill, Morgan Stanley and UBS AG."

A number of earlier articles at HingeFire talk about the Wall Street derivatives debacle:

Does the Securitization Model actually work?

The Derivative House of Cards: The “Shadow Banking” System falters

Will Someone tell the Financial Whiz Kids that their House is Built of Cards

Does the Securitization Model actually work?

The advent of securitization by banks was promoted as the beginning of a new era. Suddenly due to the mathematical modeling of Wall Street wizards, risk could be distributed while profits would be increased.

This pipe dream has come to a shattering halt over the past few months; similar to the implosion of the tech bubble in 2000. The tech industry zealously believed that standard financial ratios did not matter anymore as stocks were bid up to obscene price levels, in the same manner the financial sector put forward that the old stoic ways of doing banking were a relic of the past. At least, until the house of cards started to collapse this past summer leaving many major institutions struggling for survival.

The Financial Times discussed this spectacle in a recent article titled Payback Time.

Quants meet Reality

The Quants on Wall Street this summer received a harsh introduction to reality and the impact of multi-sigma events in the market. Traditionally, significant dislocations that tear apart the fabric of mathematically driven funds occur about every 18 years based on historical analysis. Most quants ignore history and focus on short term back-testing studies; blindly confident in their belief that these events have been arbitraged out of the market by superior math and will never occur again. Akin to the railroad engineer holding his hands over his eyes while driving the train over the cliff.

In earlier posts, the phenomena of quants was touched on:
Will Someone tell the Financial Whiz Kids that their House is Built of Cards
http://hingefire.blogspot.com/2007/10/will-someone-tell-financial-whiz-kids.html

A number of recent articles put the role of quantitative analysis and algorithmic trading in the limelight as the driver of recent market disruptions. The losses from in-house algorithmic trading desks have been extreme over the past couple of months as volatility spiked and pricing diverged from historic patterns. Morgan Stanley reported a $480 million loss in the third quarter from the bank's in-house equities trading desk that employed computer generated models to drive returns. Many other investment banks demonstrated similar issues while a number of hedge funds closed up shop.

Volatility puts algo trading under pressure
http://www.reuters.com/article/reutersEdge/idUSL2648150020071026

Nassim Nicholas Taleb discussed the impact of multi-sigma events on the market in his recent book “The Black Swan: The Impact of the Highly Improbable” “The term black swan comes from the ancient Western conception that all swans were white. In that context, a black swan was a metaphor for something that could not exist.” A Black Swan is a large-impact event that greatly deviates from the ordinary and is difficult to avoid. Taleb’s embedded thesis is that financial engineers are lulled into false complacency, not planning for the worst case and are never prepared for events that rip the fabric of their models.

The MIT Technology Review recently posted a pair of excellent articles about the role of financial engineers in the implosion of the derivatives market this past August; a crisis that is still unfolding. There finally appears to be a glimmering of understanding that the structured derivative market is a house of cards that can be crumbled by multi-sigma events, and it is just a matter of time until the Black Swan visits any leveraged market sector. The brick wall of reality trumps math every time… or in just a matter of time.

The Blow-Up: Part 1
In Wall Street's summer of scary numbers, all eyes were on the mathematically trained financial engineers known as "quants."
http://www.technologyreview.com/Biztech/19530/?a=f

The Blow-Up: Part 2
How the financial engineers known as "quants" contributed to Wall Street's summer of scary numbers.
http://www.technologyreview.com/Biztech/19531/?a=f


References:
Black Swan
http://www.investopedia.com/terms/b/blackswan.asp

Black swan theory
http://en.wikipedia.org/wiki/Black_swan_theory

Nassim Nicholas Taleb’s book can be found at:
http://www.amazon.com/Black-Swan-Impact-Highly-Improbable/dp/1400063515