fbpx

Trading with model which picks up time-series momentum. Actual trading profile for last two years has been decent with about 65% success

(Last Updated On: August 28, 2011)

I have been trading with model which picks up time-series momentum. Actual trading profile for last two years has been decent with about 65% success. The problem has been with steep drawdowns.

In some of the periods the drawdown has been severe, though it recovered fast. Have others experienced similar phenomenon where the return profile is good but drawdowns have been severe. Are there any precautions or changes to strategy which can be made to reduce the drawdown…

—–

how good are the profits?
I understand that you are getting 65% wins and rest losses/flat.. you could apply dynamic capital allocation strategies to this since its based on momentum..

Most of the time its risk vs. return and we would see sharp drawdowns for the strategies where the success ratio is higher..

—–

Exactly right – 65% I mean the same. Have used dynamic allocation based on momentum and/or volatility parameters but not been very successful. Do you recommend other ways of dynamic capital allocation.

Absolutely true – drawdown would be dependent on risk-return profile one is trying to achieve. Second is interesting that drawdown being higher with high success ratio, I have seen the opposite. Would be good to understand the situation where drawdown increases with increasing success..

 

—-

So far I am using volatility to determine my allocation and I added in profit trailing stop. This have limited my loss. Although my hit rate is about 33%, but the profit factor (or expectancy) is about 1.98. Of course, the wider the stop loss, the lesser the quantity I can trade.

—-

You do not provide enough information about the strategy, model(s) or instrument(s) traded to make much of a guess as to why you are experiencing “severe” and “steep drawdowns”…

I understand that you may not want to disclose your IP but a little more information would be helpful in the diagnosis. You may contact me privately if you are not comfortable providing any additional information in this public forum

—–

my model sifts through the intra-day prices and volume patterns to define the momentum. E.g based on last few hrs of price change and volume pattern, momentum is defined. Once the signal is given a basket of stocks is taken and held overnight and exited at fixed time the next day.
In backtest monthly profiles are good, but the issue is about some of months when there is steep drawdown of the order of 20-25%. My initial thought was that it may be caused due to some news based events but further analysis show it is not linked to some particular event. Over last 10 years of data there are two-three months every year when there are continuous losses. Few big ticket losses are followed by number of small losses.
My concern has been that in real time trading the actual drawdowns can be severe because of combined effect of draw-downs which are seen historically and also some unfavorable overnite event which might not have been hit historically.

—–

Why hold overnight? You’re exposing yourself to unexpected events.

Can you try limiting your downside with some protective calls/puts. You can sell them back at the open to prevent the risk of overnight events.

 

– Positions can be liquidated before close but that that penalizes the profitability to a great extent.

Protective calls/puts help in reduction of drawdown but penalizes the Sharpe Ratio.

—-
Depending on your frequency of trades, setting up control system might be useful for example quadratic loss function, also there are number of pattern recognition tools(data mining) that might solve the problem

Thanks

—–

I don’t do this myself, but I’ve heard of some traders who stop trading for the month if their losses exceed a certain % of their trading account. Perhaps you could examine your loss profile and determine a threshold beyond which you know you’re going to experience “one of those months” and just stay out of the markets once the threshold triggers.

Also, you could try a profit-taking stop as well as a trailing stop loss. The profit-taking stop will typically increase your winning % while lowering your net profit somewhat, but it may also reduce your exposure by getting you out of the market sooner. The trailing stop loss will usually reduce your winning % while increasing your win/loss ratio. It may also impact your net profit, but also serves to reduce your time in market.

 

 

NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!

Subscribe For Latest Updates

Sign up to best of business news, informed analysis and opinions on what matters to you.
Invalid email address
We promise not to spam you. You can unsubscribe at any time.

NOTE!

Check NEW site on stock forex and ETF analysis and automation

Scroll to Top