Snapman
Posts: 624
Join date: 2009-06-25
Age: 24
Location: New York City
 |  Subject: Excellent Article about effects of Quant related HF's by - Alex Burns of Telesis Capital, LLC Tue Dec 01, 2009 8:02 pm | |
|
|
|
|
|
|
|
|
|
| | Click Here To Read Comments on This Story or Submit Your Own | ![]() | Guest Article: A Quant
Bubble? More Likely a Mental Bubble |
Over the past several months, there has been a preponderance of
negative press regarding the market influence of quantitative trading
strategies. Quantitative high frequency trading, and more recently,
momentum-based quantitative trading strategies have suffered the brunt
of these attacks. A recent article from Doug Kass appeared on the
TheStreet.com contending the latest growth in global asset values is
largely due to quantitative momentum-based and high frequency trading
strategies. This argument fails to hold water on any
level. The assumptions are wrong, the logic is disconnected, and the
conclusions are completely invalid.
First, let's take a big step back and understand what quants actually
do. The majority of quants employ some subset of five strategies. They
are Trend, Reversion, Yield (aka Value), Growth, and Quality.
Furthermore, quants implement these strategies over a variety of time
horizons, from high frequency (intraday) to long term (a year or even
longer). Quants trade in a variety of liquid markets, including stocks,
futures, foreign exchange, and, to a small extent, options. Quants
either trade directionally or on a relative basis. Relative trading
involves betting that one kind of instrument will outperform another
without knowing whether the instrument will actually go up or down on
its own.
Some basic examples: equity statistical arbitrage ("stat arb" for
short) usually employs mean reversion on a short term time horizon in
equity markets, and it makes relative forecasts. In an applied setting,
stat arb traders are often long the underperformers and short the
out-performers within a given sector. Quant Long/Short ("QLS" for
short) strategies generally utilize relative bets in equities based on
the more fundamental strategies listed above, yield, growth, and
quality. QLS strategies, like stat arbs, are most often implemented on
a market neutral basis. High Frequency Trading ("HFT") can be done in
any type of liquid market, and most often involves mean reversion (or a
close cousin, market making). HFT can be implemented either in a
directional or a relative basis, but the
majority of HFT activity tends to "go home flat," which means that
these traders do not keep positions overnight. Whatever they buy is
sold by the end of the day (or vice versa). Futures trend following is
another example of a popular quant strategy. These are trend strategies
implemented in futures and FX markets, most often on a long-term time
horizon, and they are directional. This means that they are making
outright bets on each market.
It is easy to demonstrate that those implementing "relative"
strategies and those implementing HFT strategies are unlikely to be
among the ranks of bubble-causers. Kass' argument essentially equates
quant trading with momentum trading, claiming that some two-thirds of
all volumes are being driven by momentum traders.
Two problems here.
First, the Aite Group study Kass cites for his volume figure presents
this statistic for individual U.S. equities, and it does not
distinguish between the myriad sources of algorithmic trading. The high
frequency traders that are behind the bulk of these volumes are useful
to the markets. As a direct result of the liquidity they provide,
bid/ask spreads across the equity market have declined dramatically
over the past several years, making these strategies important to
smoothly functioning markets. These traders do not take positions
overnight. This makes it hard for them to drive bubbles, given they
aren't accumulating positions for any reasonable length of time.
The second problem with Kass' argument is that a large
proportion of all quant trading in equities actually bets against the
prevailing trend, either by providing liquidity (making markets), or
trading a mean reversion (stat arb) strategy. In equity trading, these
strategies generally create market neutral portfolios, betting that
spreads between stocks reverse and/or that the performance of a single
stock reverses.
Even for those quants employing momentum strategies in stocks, their
portfolios would still generally be market neutral, which means they
are both long and short roughly equivalent amounts of stock. Such
traders, even if they are implementing momentum strategies, are not
moving the equity index levels as a result of their actions. The
massive quant deleveraging event of August 2007 provides compelling
evidence: the S&P500 was basically unchanged while quants
liquidated nearly 70% of their positions in about two weeks and this
was when quants held substantially larger positions than they do now.
Not all momentum traders are quants either. Many discretionary
investors, from pensions to global macro funds to moms-and-pops,
believe in buying whatever's been rising and selling whatever's been
falling, which is straightforward momentum trading. Even if quants
account for two-thirds of all equity volumes, this is demonstrably
irrelevant to the forming of a bubble in equities, much less in gold or
other asset classes.
In theory, longer-term trend following strategies could cause
a bubble in asset values, but these strategies would have to be
directional, and they'd have to dominate the trading volumes of their
markets. We know the majority of directional momentum strategies in
quant trading are deployed by CTAs in futures markets. The same Aite
Group study already mentioned estimates 30% of futures, and about 25%
of FX volumes, are driven by HFT players, hardly a domination. More
importantly, we have already dismissed HFT as a driver for an asset
bubble (because bubbles require accumulation of positions), and I have
seen no evidence of the proportion of volumes accounted for by
directional quants generally. As such, we have to look at this argument
other ways. According to HedgeFund.net, CTAs
currently manage approximately $170 billion, only a portion of which
would be in systematic long-term trend-following strategies. By
contrast, The Canadian Pension Plan Investment Board manages about $130
billion, and that's one pension fund. In Canada. CalPERS weighs in at
around $200 billion. So, not only are CTAs not very large in the scheme
of things, but they are also usually diversified among the market they
trade, often splitting up risk among 50 or more different futures
markets. Additionally, these strategies are long term, with average
holding periods on the order of six months, meaning they turn over only
a small portion of their positions every day.
Kass concludes with an example of the heady compensation
available to high frequency quants. Here, he actually has a fair and
valid point. I too believe there is a bubble in quant-land, but it
relates to the compensation of practitioners, and to the valuations
assigned to their companies, not to any interaction they have with the
capital markets. That a smart UPenn grad can make $200k right away does
not imply that quants have driven gold to $1150 an ounce. These are not
related statements.
I get it: this steamrolling bull market doesn't seem to make a
lot of fundamental sense to me either. Kass is angry and looking for
someone to blame. The $54 billion invested into 22 commodity ETFs
(according to State Street) seem a plausible candidate for having
produced a commodity bubble. It's also possible that this isn't a
bubble at all. But the idea that quants are driving markets is, in
short, foolishness of the highest order.
Rishi K Narang is the Founding Principal of Telesis Capital, LLC, which
invests in quantitative trading strategies via separately managed
accounts. He has over 13 years' experience in hedge funds, most of it
as a practitioner of and/or an allocator to quant trading strategies.
Narang is the author of Inside the Black Box: The Simple Truth About Quantitative Trading (www.thequantbook.com). Telesis can be reached at query@telesiscapital.com.
Alex Burns is in charge of due diligence at Telesis Capital, LLC. He
is a CFA charterholder, and has been analyzing quantitative hedge funds
for over eight years.
|
|
|
Snapman
Posts: 624
Join date: 2009-06-25
Age: 24
Location: New York City
| | Subject: Re: Excellent Article about effects of Quant related HF's by - Alex Burns of Telesis Capital, LLC Tue Dec 01, 2009 8:05 pm | |
| The point is Quant Funds provide great liquidity for the small guys like us and for market markets (ibanks). These quant funds are good for the markets but are not fueling the excessive speculation. As the article points long directional strategies probably are what cause bubbles to occur. And they would have to be a consideralbe amount of the market in which they are bubbling.
The only thing I can think of so big that are long and directional are mutual funds. I know batman and sauros have some great research on this! Love to hear your thoughts on what you guys found out here.
-snapman |
|
