Are Quant Funds The New Chart Traders?
Do new and sophisticated quant funds have better chances to profit than the chart trader of the 1990s that used technical analysis? Is it possible that quant funds are the equivalent of retail traders of the past?
The retail chart traders of the 80s and 90s were easy targets for systematic trend-followers. After all, where did the high returns of CTAs come from during that period? Trading, and especially futures trading, is a zero-sum game. CTAs by being patient trend-followers took advantage of the random methods of chart traders and profited at their expense. Some that are new to trading are not aware of the frenzy in the 90s about intraday trading mainly financed by brokers. Systematic traders took advantage of it and made large returns. But after the random intraday and swing traders were driven out of the market, CTAs have had problems generating returns. There is scarcity of retail dumb money at this point. Neal Berger, the CIO of Eagle’s View Asset Management thinks that “pedestrian” quant strategies will not survive:
“Systematic strategies require an endless supply of victims to thrive, and the growth of quant and passive funds has caused dumb money to behave unpredictably or disappear altogether. “ Source: Bloomberg
Chart traders of the 80s and 90s used a mostly random trading method called technical analysis that facilitated wealth redistribution from its naive users to professionals. Are the methods quant funds using nowadays less random or better than the technical analysis of the past century?
When looking at the objectives of some quantitative funds about the strategies they are looking for one may wonder if these quants have any experience with trading and especially of the systematic type. A high number of requirements translate directly or indirectly into more parameters, mathematical or qualitative. More parameters mean higher chances of over-fitting to market conditions that may disappear and not repeated in the future.
Quant funds must understand that the market is not a money machine that prints on demand. Making money means some other traders must lose money. At this point the market is driven by algos to the tune of 80% of the daily volume. Where do those quant funds plan to extract alpha from?
A retail quant trader or a small quant fund ($10 – $20 million) that uses idiosyncratic alpha strategies has an advantage and can exploit price action inefficiencies without causing adverse price moves in higher liquidity markets. But a large quant fund does not have the luxury of exploiting anomalies and must find strategies that are less idiosyncratic in the way they generate alpha, for example market neutral equity long-short or cross sectional momentum.
The challenge is how to develop a strategy for high capacity without risking becoming the target of predator algos. The usual way is by considering a large universe of securities and diversifying across sectors to keep beta low while maximizing alpha. This sounds good in theory but is it fundamentally different from what all those failed chart traders of the past century did? Or to put the question on more practical grounds: are those intense efforts by quant funds using machine learning, data science and advance programming languages going to generate a result better than the long-only moving average cross applied to SPY ETF since inception, as shown below?
Figure 1: Backtested performance of long-only 50/200 moving average cross in SPY from inception to 08/31/2018.
The above backtest includes $0.01 per share commission and equity is fully invested. CAGR for the strategy is 9.85% versus 9.68% for buy and hold but risk-adjusted performance is even higher with MAR (CAGR/maximum drawdown) for the strategy at 0.50 versus 0.18 for buy and hold. This is the result of a much lower drawdown of the strategy due to avoidance of bear markets. But will the strategy manage to avoid future bear markets before generating high drawdown? This will depend on the speed of falling price action. Up to this point, this strategy, which is probably the result of data-snooping (we would not be talking about it if it had not survived,) has managed to avoid large drawdown but a Monte Carlo simulation of equity curve changes (not the trades) shows that the probability of a drawdown greater than 38% if about 10%. In other words, it is risky to invest a lot of money in this strategy without proper hedging.
Figure 2: Monte Carlo simulation of maximum drawdown for long-only 50/200 moving average cross strategy in SPY
I highly doubt most quant hedge funds will find anything better than the above moving average cross no matter how many PhDs they employ and how many fancy presentations they make. Probably a few small niche funds will find ways to generate alpha by staying under the radar (this is an actual expression used by a quant fund manager who is using our software to generate features for one of his strategies.) The quant funds that are misguided either by lack of market experience, or due to wishful thinking or even by relying on academics who look at trading from different angles, will probably have the same future as the chart traders of late last century. During those times it was the retail chart trader against the systematic trend-follower and the specialist or market maker in the futures pit. Nowadays, it is the aspiring quant fund and systematic traders against algos and massive passive funds. And you may suspect who the winners will be.
In my opinion the future of the quant fund is the small operation ($10 – $20 million) with minimum expenses. The fund employs idiosyncratic alpha strategies and shares profits with a few investors. For a fund with $20 million AUM, 10% return translates to $2 million of profit. A 50% performance fee (no management fee) leaves $1 million to cover expenses and compensation. Usually the fund may employ up to two programmers and salaries plus bonus will be about $200K per year. Operational expenses may run up to $500K and that leaves a decent salary for the fund manager but nowhere close to what people made in the 90s. We should forget the 90s and constant reference to that period serves no useful purpose any longer as the dynamics of markets have changed. If anyone believes there is substantial alpha to be made in the markets they should point to its source. Large quant funds will face severe problems and maybe losses from constant friction generated by algos and other market participants, including company insiders and analysts.
About the author
Michael Harris is a trader and best selling author. He is also the developer of the first commercial software for identifying parameter-less patterns and related anomalies in price action 17 years ago. In the last seven years he has worked on the development of DLPAL, a software program that can be used to identify short-term anomalies in market data for use with fixed and machine learning models. Click here for more.
No part of this article constitutes an offer to buy or sell any sort of security or financial product. Furthermore, no part of this article should be construed as advice or a recommendation for any type of financial transaction or investment decision. For our full disclaimer click here.
© 2018 Michael Harris. All Rights Reserved
Risk Warning: The FXCM Group does not guarantee accuracy and will not accept liability for any loss or damage which arise directly or indirectly from use of or reliance on information contained within the webinars. The FXCM Group may provide general commentary which is not intended as investment advice and must not be construed as such. FX/CFD trading carries a risk of losses in excess of your deposited funds and may not be suitable for all investors. Please ensure that you fully understand the risks involved.