Approaches: Challenges and Remedies
Posted by orangequant on December 29, 2006
Orangequant Comments:
it might not be obvious to you when you first visit this blog just how complex an undertaking this is. when we start posting our activity next week all you’ll often see is the trades and gains/losses. but there are underlying issues as well- things which make this 10% compounding goal very challenging…
Scmfinance and i (orangequant) will vary in our techniques for handling these challenges, i’m sure. scm has already mentioned these issues (“…settlement periods, commissions, and trading on multiple boards…”) and i’d just like to elaborate a bit and share my own plans for coping with these challenges. (Scm recently mentioned to me the idea of “two-tracking” our initial bankrolls and i hope scm will fill us in on that some more, as i think it’s very important.)
the 3-day settlement rule actually has three edges. first, it requires the trader to wait three days after closing a trade for settlement and availability of funds; second, if the underlying security has already been held for at least three days, then the close produces “instant cash” (except as in the following third edge case); third, if the trade is closed at any time before the lapse of three days from its open, then the 3-day rule engages afresh (starts the three-day count over). instead of “sell”, i use the phrase “closing a trade” here so as to include both long and short trading.
<>this third edge of the 3-day rule can deal a crippling blow to any plans for minimum once-a-week roundtrip trading. Ideally, to maximize the periodic number of trades under the 3-day rule, one would want to exit the trade either on Day Zero (i.e., same day roundtrip), or on Day Three.
The Tortoise and The Hare
of course, the Elephant In the Room here is gain maximization per trade vs long-term gain maximization. in other words, will you do better to focus on maximizing your gain in each trade and ignore the 3-day rule, or will you do better to focus on maximizing the number of trades over the year? instinct and “common sense” tell us to maximize gains on each trade. that’s how most traders behave and, in my opinion, it’s dead wrong. instinct and common sense are just no match for spreadsheet wisdom. across one year, assuming equal skills, the trader focused on maximizing the number of trades will do very substantially better than the one focused on maximizing per-trade gain.
i’ll provide some charts here later to illustrate my point, but for now i’ll just make the point that compounded gain is strongly frequency-dependent. since about half the time my approach is through quantitative analysis (“quant”), i try to notice these little things. in all of this it’s also important to realize that the 3-day rule, used efficiently, actually permits the closing of up to 1.66 trades per week (after the first week).
<> Reader comments are welcome. Scm, what do you think?
scmfinance said
as orangequant and i have both pointed out, the challenges facing this experiment are ridiculous — almost in a laughable way. settlement periods are always the most frustrating for me and considering we’re only using $500 that means we won’t have the luxury of a margin acocunt. orangequant has already defined the 3-day settlement rule, so there’s no need for me to do so again. it’s extremely frustrating because i can’t tell you the number of times i have purchased a stock, sold, wanted to enter another play using those funds (which were unsettled), but chose not to enter because i wouldn’t be able to exit and didn’t know if the play would hold the gains long enough for the funds i used to settle.
i may experiment using a “two-track system” where i divide the money in half so that i can move more fluidly in and out of plays. but that reveals a couple more problems: a) i’ll have to make twice as many picks (not easy) and b) my returns at the beginning will be smaller which on a 10% gain will make paying the commissions much more difficult. so, if i do utilize a “two-track system” it will most likely not be at the beginning of the experiment.
while neither strategy (maximizing gains on fewer trades nor frequent trading with smaller gains) is easy, i personally think that frequent trading with smaller gains is the better route to go. the bottom line is that when the smaller gains start compounding, they add up and yield nice annual gains — or at least that is what i hope to show. the 3-day rule may cause this strategy to start off slowly, but as orangequant has shown, 10-20% really starts to take-off.
How to Get Six Pack Fast said
I noticed that this is not the first time at all that you mention the topic. Why have you decided to write about it again?