Posts tagged ‘Backtesting’
System Lab: March 2012 Performance
The system closed the month of March with a modest gain (+2.5%). With the exception of a few days, most of the market’s movement continues to be in the gaps, with relatively narrow intraday ranges. Again and again the market gaps up/down and then just lays there. Everyone seems to be guarding their chips and waiting for the other guy to make the first move.
So why all the caution? My personal view is that the ever-looming Euro debt crisis has instilled quite a bit of fear into the markets and the retail investor (after loosing all of his home equity and most of his 401K) has run screaming, fleeing equities for the foreseeable future. Thus, what we seem to have now is a listless market comprised largely of HFT bots trading money back and forth. A zero sum game and a difficult environment for the retail day trader, to be sure.
What Happened To The Quants In August 2007? (paper)
Interesting paper that discusses outlier losses experienced by long/short hedge funds during 2007 and proposes an interesting EOD RTM strategy (one I developed and tested myself prior to coming across this paper, dammit!):
“Given a collection of N securities, consider a long/short market-neutral equity strategy consisting of an equal dollar amount of long and short positions, where at each rebalancing interval, the long positions are made up of losers (underperforming stocks, relative to some market average) and the short positions are made up of winners (outperforming stocks, relative to the same market average). By buying yesterday’s losers and selling yesterday’s winners at each date, such a strategy actively bets on mean reversion across all N stocks, profi ting from reversals that occur within the rebalancing interval.”
(click on title for link)
System Lab: January 2012 Performance
January was a fairly moderate month and its limited volatility offered pretty slim pickings. The system had some nice returns right out of the gate, but these were pared back as the month progressed and most days were under +/- 1%. The good news is that the system was still able squeeze out a modest gain of +1.8% for the month. Not earth shattering, but I’ll take it.
Three Consecutive Higher Closes
This system trades a variation of the “3 higher closes” strategy on the SPY index ETF. It trades from the long side only, using the following entry criteria:
Buy SPY at the close if
- It has closed higher three consecutive days in a row and each close > open.
- SMA50 > Close > SMA200.
- Yesterday’s volume > SMA50
The system holds for N days and exits at the close. Here’s a chart showing the reward/risk ratio vs hold time:
Some caveats include that backtest only generated 6 trades in over 10 years and the results didn’t include slippage/commission. Hardly statistically significant, but interesting nevertheless…
Strong Closes for Modest Gains
Here’s a quick little study on a simple mean reversion technique using the SPY ETF:
…when SPY closes strong (in the top 10% of its range) but still only manages a small gain on the day, that the next day has a downside tendency
I suspect that when you factor in commissions and the fact that one cannot conditionally short the close price of any stock (MOC orders generally have to be entered >10min prior to the close), the edge might prove artificial. Interesting food for thought, though…
Nuts & Bolts: Survivorship Bias
Survivorship Bias (SB) is a plague to most trading system developers. Pick any basket of stocks to trade and you’ve just introduced it into your results, especially if that basket is a collection of index components (like the S&P500, Nasdaq 100, etc).
So what exactly is this bane of all algo traders? Here’s the textbook definition:
In finance, survivorship bias is the tendency for failed companies to be excluded from performance studies because they no longer exist. It often causes the results of studies to skew higher because only companies which were successful enough to survive until the end of the period are included
Global Optimization Algorithms: Theory and Application (e-book)
Well written e-book on various optimization strategies using bio-inspired/evolutionary techniques (click title).
Risk and Return in Momentum Strategies (paper)
An interesting paper on the classic Momentum strategy but with a few new twists to factor in tail risk (click on title):
“We apply risk-return ratios at the individual security level in order to drive the stock ranking process (construction of momentum portfolio) and at the portfolio level in order to evaluate and optimize the risk-return profile of the winner and loser portfolio. We investigate whether the application of risk-adjusted criteria with balanced risk-return performance can generate more profitable strategies than those based on a simple cumulative return criterion which serves as a benchmark. Moreover, by introducing risk return ratios as portfolio selection criteria, we are able to postulate a portfolio optimization problem with a ratio as an objective function. Therefore we devise an optimized-weighted strategy that creates optimal risky winner and loser portfolios”
Bootstrapping: White’s Reality Check
Interesting article on WRC (click title):
“Prior to WRC, bootstrapping could be used to generate the sampling distribution to test the significance of a single rule. White’s innovation, for which he was granted a patent, allows the bootstrap to be applied to the best rule found by data mining. Specifically, WRC permits the data miner to develop the sampling distribution for the best of N-rules, where N is the number of rules tested, under the assumption that all of the rules have expected returns of zero. In other words, WRC generates the sampling distribution to test the null hypothesis that all the rules examined during data mining have expected returns of zero.”
Bleed or Blowup? Why Do We Prefer Asymmetric Payoffs?
As a trader which would you prefer: A slow bleed of small losses with occasional big winners (ala momentum strategies) or large infrequent losses with many small winners (ala mean reversion)? This interesting 2004 paper (by none other than Taleb himself) discusses.