New Members Entry Point – If you’ve arrived on this page looking for the k1 Project and all the reference material on Phil’s strategy, follow this: Front Page of the K1 Project.
k1p Table of Contents
- Introduction – motivations, approach, and rules for the virtual portfolio
- Supporting Analysis – A Year of SPY Trades (this page)
- The Virtual Portfolio – positions and actions
As a part of my research into the LTP method, and my recent Aha moment about managing delta, I decided to do a little backtesting of the strategy I’ve defined. Thanks to the thinkback feature on ToS, I was able to pull up the closing prices for all of the options chains for SPY over the past year, and work out a pretty simplistic set of trades I could have followed through that time period.
No market timing, no uncovering, no time spent in front of a computer screen. For the most part, I simply waited until expiration day and rolled the short caller forward. A couple of times the market dropped significantly during the period, and so I rolled down both my long and short sides, but even that is pretty much automatic. There were definitely opportunities to sell an additional call strike on the way down, but I didn’t bother optimizing the results for that. I also didn’t nitpick at the spread the way I would do in actual trading, so there are at least a half dozen nickels available in there as well.
The results? About 36% over the past year, from just after June07 expiration until today. Including all of the market meltdowns, the rallies, the craziness. Using my ToS commission schedule, I count 38 trades so far, for a grand total of $57 in commissions. I have the results in a spreadsheet, but every attempt I’ve made to enter them here does not work. So you’ll need to download the spreadsheet to look at it.
What this means to me is that the hedging strategy can work just fine over time, and we can remain market neutral the entire time. This has been a terrible year so far for SPY, and it shows. The strategy is down about 10% from Christmas, although it’s hanging in there.
Bronek has performed his own backtest of the historical trades, and provided PDF files showing the results of the trades. I’m posting them here, with Bronek’s descriptions:
Ref 1. (Link) In this time frame market (SPY) declined from ~153 to ~132. Entry cost was ~$2060 trade ended with ~-365 cost i.e. ~$1700 was recaptured, and the value on 7/03/08 was $1195 ($830 profit). It is not worth of my time to calculate P/L b/c of declining value of the initial outlay. The risk graph is a std risk graph, flipped and rotated so the price axis is aligned with adjacent stock price. Therefore effect of the price movement on the P/L is easily visible. The colored lines on the risk graph represent the trade values if the price would have reminded constant and only time would approximate the expiration. Other two graphs represent historical volatility, and effect of volatility crush.
Ref 2. (Link) In this time frame market rose form ~126 to ~153. 126 is close enough to 132, for nongovernmental work, to consider that evaluation represents a round trip up and down. The trade was opened with ~$1552, additional ~$1000 was required during this time to pay for rolls. The trade was close for $297 profit.
Ref 3. (Link) Combined trade 1 and trade 2. Result is: all but $68 was recouped by positive cash flow, and the trade if liquated would have provided $1127 profit i.e. above $68.
