Optimal f vs. Margin

[arold_ite]

As part of my trading strategy design, I’ve spent the last month researching specifically Money Management (position sizing, bet size, call it what you will).

Having read three (very deep) books and countless articles on the web, as well as running numerous simulations myself, I have come to the conclusion that searching for the ‘Optimal f’ (i.e. what is the correct fraction of equity to risk per trade) is a useless endeavour (in my case).

Why? :- MARGIN

As an example, if I calculate my optimal fixed fraction of equity per trade to be, say, 2%, this means that on a starting account of $10,000, I should risk $200 per trade.

OK, so let’s say that I’m trading the mini-sized Dow ($5 per contract), which has an intraday, initial margin requirement of $1,250.

Let’s say that my strategy calls for a tight-stop loss of just two points. This means that I should open a position with 20 contracts, (200/5)/2

So if I lose this trade, I will only lose 2% of my equity (20 contracts multiplied by $5 multiplied by 2 points = $200).

So far so good, I have a fixed fractional risk of 2%, I have a fixed stop-loss, which is based on the dynamics of the market, not how much I can afford to lose and I have calculated how many contracts I should order (20).

But (and here’s the whole point of my post), because I have $10,000 net liquidity and the intraday initial margin on this contract is $1,250, how many contracts can I order?

Well, it’s not 20, it is in fact, 8 !!!

By ordering just 8 contracts against an account with $10,000 liquidity, and a stop-loss of two points, instead of risking my desired 2%, I’m only risking 0.8% (8 contracts multiplied by $5 multiplied by 2 points = $80).

You might think that this is a good thing because I’m not risking so much. But, the whole point of going through the process to discover the optimal fraction is to MAXIMISE my profit AS WELL AS minimise my risk.

You might also say that if I can’t order so many contracts because of the margin, then I should widen the stop-loss in order to bring the 0.8% risk up to my desired 2% risk but that would be foolish because the stop-loss should be a calculation of your system (i.e. a reflection on the price/indicator/volatility/support/resistance or whatever you happen to be using) and NOT a reflection of how much you can afford to lose. The market doesn’t care how much you can afford to lose.

The bottom line here (as I understand it) is that; all of the research and work to find the optimal fraction of equity to risk, isn’t worth a hoot because I can never trade the desired number of contracts as I am limited by MARGIN.

I’ve also come to the same conclusion with many other contracts that I plan to trade, not just the mini-sized Dow. For example, the FTSE 100 allows a maximum fraction of 1.06%, the DAX is just 0.44% and the EURO FX is 1.54%.

Of course, it is always possible (some would say probable!) that I have completely misread the situation and have just made a fool of myself with these assumptions.

Actually, I really hope this is the case and that somebody can tell me that I’ve miscalculated these assumptions.

Until then, I will have to rethink my money management strategy to take margin into account when calculating the optimal fraction of equity to risk.

Would anybody care to comment on this?

Arold

[spike500]

arold_ite

The margin you are talking about is to go overnight.
Intraday the margin will be about 500$.

If you go overnight, i think a 2 points stop might be too close.
Which means that your risk will get higher. If you put a stop on 3 points the risk will be 50% higher.
You also have to take in account the commissions you pay and the slippage. All these elements add up to the loss. So your loss will never be 2 points or 10$ but more.

The margins are there because they are necessary. The margins are the only element in your calculations that you can't change. They are imposed.

Regards

[arold_ite]

Thanks for the reply.

Take a look at this link:

http://www.cbot.com/cbot/pub/page/0,3181,1041,00.html#1d

Which shows that the Initial margin on the mini-sized DOW is actually $2,500. Yes, this is for overnight maintenance. But, my broker (IB) accepts 50% margin for same-day trades up until about 15 minutes before the close. This is why for me, intraday margin IS $1,250 and as such restricts my risk profile.

I here and understand what you say about overnight positions, etc. The point I'm trying to get at, is that 'standard' 'Optimal f' calculations are useless for day-trading futures.

Or am I wrong?

Arold

[spike500]

intraday margin is only 500$ with my broker ( and with most brokers).

so you need only 500$ instead of 1250$.

[arold_ite]

$500 intraday would just about get around my problem.

My problem is that I'm developing an Automated Trading System and Interactive Brokers is the only company I've found so far that can support an ATS and also allow me to trade in the number of markets that I want to.

I wonder why their margin requirements are so high?

Arold

[soylent]

arold --

There are other brokers w/ automated trading. Just check them out / look around (MAN financial and others).

BUT: The more important thing:
I think you are setting an entirely too-tight stop on YM.

Why? Because:
a) If you really think you'll sell @ 2pt stop, then you need to sell @ a 1 pt stop to account for the bid/ask spread. Otherwise, you'd have to set a 3pt stop and 'execute' the stop @ 2pts stop.

b) YM moves 5-10 ticks pretty fluidly. Even on 3-min bars, those oscillations aren't unheard of.

Realistically, on YM, you'd need 11pts-ish (assuming you're taking signals off 3m bars) to allow your plan to develop. YM spreads frequently open to 2-3pts. Granted you don't have to go 11pts -- but, I find, most the scalpers stop out by 8-9pts. I keep mine well away from there (20pts) and tighten if my signal reverses before the market does -- but, I trade larger time frames.

-c

[arold_ite]

I take your point about the stops soylent.

To be honest, I was using 2 points as an example. My trading system uses recent volatility to set stops and it would be very rare to have a 2 point stop.

Cheers
Arold