Tortoise and the hare

[rathcoole_exile]

Indeed. In all truthfulness, I chose 20% for a specific reason. I'm starting out with an older system (written and last modified in late 2004) that has shown fairly stable returns over time. This provides a prety good set of data for estimating optimal f, which turns out to be 24% over that time period. I figure 20% should keep us a bit below opt f to help out with short-term variations in the underlying return distribution while still being close enough to give the optimal f theory a fair go in real life. That is the most interesting part of this exercise to me.

My expectations are for the account to basically spike up and down between 2k and 20k for a period of months, always reverting fairly quickly to about 5-10k. This behavior is caused by short run-ups followed by shallow drawdowns If we're lucky, the system will catch a long enough run at some point before the account heads down to zero, and the equity curve will "break out" in a big spike. I have yet to determine specifically where to start dropping leverage (toying with fixed ratio), but I've got some time to figure that out as all reinvestment methods behave about the same when the account is small like this.

As I said, this account is pure risk capital -- an amount that is unimportant to me. I expect to learn a lot about risk and my attitude towards risk regardless of the outcome, and the risk of losing the capital in the account is a small price to pay for this knowledge.

jj

JJ, sound, logical arguments and you're certainly going into this with your eyes wide open about the potential risk. I hope it pays off for you !

 

[Mathemagician]

Just a quick update... The account that I'm running as part of this experiment just broke through the first milestone ($5k) and will trade 2 contracts today. This is where things can get ugly, as the account struggles mightily to work its way above 5k on 1 contract before being slapped back down hard on 2 contracts, only to start the struggle all over again with 1 contract. This can go on for weeks/months and is a spot where these projects have a rather high mortality rate. If we're lucky, it'll break through at some point with a healthy run. If not, then we'll eventually hit a drawdown and get dinged out.

jj

 

[rathcoole_exile]

Just a quick update... The account that I'm running as part of this experiment just broke through the first milestone ($5k) and will trade 2 contracts today. This is where things can get ugly, as the account struggles mightily to work its way above 5k on 1 contract before being slapped back down hard on 2 contracts, only to start the struggle all over again with 1 contract. This can go on for weeks/months and is a spot where these projects have a rather high mortality rate. If we're lucky, it'll break through at some point with a healthy run. If not, then we'll eventually hit a drawdown and get dinged out.

jj

interesting report; please keep the updates coming ......

 

[Mathemagician]

interesting report; please keep the updates coming ......

Thanks! Will do. FWIW, I did start a blog for this exercise that follows the day-to-day in excruciating detail.

jj

 

[Mathemagician]

Just a quick update to let you know that the experimental blog account has finally reached the point where it is doing some position sizing. We traded 2 contracts for the first time (first time in a while, anyway) yesterday and this marks the beginning of the interesting part of the experiment.

jj