I have also been tracking a strategy which mock trades a portfolio of 10 large cap individual stocks. I added a 7th stock in the last week of Dec. 2003 and 8th and 9th in January '04, and a 10th in late June '04. These last 4 were stocks which I liked early on, but their prices post-crash had become too low to allow easy shorting. Since their prices have recovered I've added them to my list. Note that a longer list will allow a larger portfolio to be traded without overtaxing the depth of the market at turning points, so I may continue from time to time to search for stocks to add to the portfolio. The strategy is independent of whether the stocks are trending up or down on the intermediate or long term. The charts below show the net portfolio value including slippage, spreads, and TradeStation Securities commissions since May '02. While the stocks were chosen more or less at random - big cap NASDQ stocks being the only a priori requirement, there is some survivorship bias in the results - the original incarnation followed 9 stocks, and the worst performing 3 stocks are no longer followed. However, the best stock I've found for this strategy performed much, better than these 9, and it too is not included in the performance chart (but I traded it often with real money. Alas, the company was bought out in early '04). No stock ever examined with this strategy has produced a negative return for longer than a few months, and even then by only a very small amount. There's an important caveat: in some cases a given trade might not have been filled if the depth of the market at the limit price was insufficient. Using TradeStation Securities or another broker and trading software which can mask one's trades against NASDQ Level II visibility will help significantly. Even so, the larger the value of the portfolio, the more this will be a big problem. Practically, one would have to either expand the number of different stocks traded as the portfolio grew, or pull the cash generated out of the account and invest it elsewhere. The performance chart should thus be interpretted to demonstrate the potential of the model to generate cash when the depth of the market accomodates. Accounts up to ~$100,000 or so should be fine with these stocks and strategy. As you see below, the returns are strong and remarkably free of volatility. This model has also been appreciating at over 100% per year average, and with remarkably little draw down or volatility. Initial invested stake is assumed $10,000 per stock; includes TradeStation Securities commissions on all trades. Prior to Feb 27, I had plotted using $20,000 per stock initial invested. The lower initial stake is probably more appropriate - that's about $70,000 total initial portfolio. With the lower $10k initial stake commissions are a larger fraction of the profit per trade, lowering returns from +145% per year to +137% per year as calculated on Feb 27, 2004.
As of June '05, I have discontinued following this model. I've had too many experiences trading stocks which lead me to think that the depth of the market required for this strategy is higher than I have been assuming. Add to this the eery lack of volatility in this relentlessly appreciating model at 100%/yr. Seems hard to believe that this could not have already been discovered and exploited by hedge funds to an extent at least sufficient to reduce the return substantially over the time period I've followed it. I still think this is a good strategy, but probably only for a small part of any portfolio; a few hundred shares maximum per trade. I use the basic idea often in setting up limit orders with my actual trading, usually to good outcome.
