God bless IB, but you won't get rich with IB

[Businessman]

Quote from whitemills:

I don't view this as a hateful quote at all... I think it's a very safe bet that if the output of your position sizing technique ever conflicts with the minimum margin requirements of IB, your risk of ruin is definitely close to 100%... regardless of whether you use IB or go for the $500 margin guy. It's just a matter of how long it takes you to run it into the ground.

Can anyone most the math that says otherwise, without winning on 90% of your trades?

Not completely true. If you day trade a basket of diverse futures and allocate 20% of margin for each instrument and use stops on so your risk is just 2% on each. Your risk of ruin is very small but a 100% margin increase will force you to cut size below what you would like.

This is what is happening to me.

 

[whitemills]

Quote from Businessman:

Not completely true. If you day trade a basket of diverse futures and allocate 20% of margin for each instrument and use stops on so your risk is just 2% on each. Your risk of ruin is very small but a 100% margin increase will force you to cut size below what you would like.

This is what is happening to me.

So as an example, let's say you have $20K in your account and you allocate 20% to NQ futures, which have a margin requirement of $4K. You can trade one contract. Your risk level at 2% is then $80 which equals 4 pts on that contact. Let's say you traded someplace which let's you trade 2 contacts under your system - now at a 2% risk, you have to raise your stops to 2 pts away from your entry.

Is your system that accurate or did I err in my analysis above?

 

[Businessman]

Quote from whitemills:

So as an example, let's say you have $20K in your account and you allocate 20% to NQ futures, which have a margin requirement of $4K. You can trade one contract. Your risk level at 2% is then $80 which equals 4 pts on that contact. Let's say you traded someplace which let's you trade 2 contacts under your system - now at a 2% risk, you have to raise your stops to 2 pts away from your entry.

Is your system that accurate or did I err in my analysis above?

More like this:

Lets say i have 200K, so i allocate 20% to each instrument: i can trade 20 contracts of NQ if there is 50% intra day margin.

My desired risk is 2% of my account per trade which is 4K to stops ie 10 NQ pts on 20 contracts = 4K stop loss. I use similar sizing for CL, GC, EUR etc

Now margin are raised by 100%.

I still have more than enough margin after the increase to trade 20 NQs on their own. But now i cant trade all other makets at the same time.

So now i have to cut my size down to 10 NQ contracts instead of 20 contracts if i want to use the same 10 pt exit and day trade the other markets too.

 

[whitemills]

So you are obviously trading an intraday, very short-term system. Does your system typically have you in all five markets at the same time?

I'm just trying to probe you on your position sizing... Its seems like you are basing your stops on your position size and the maximum number of contracts allowed by the margin rules.

Just to be clear - you are stating that when you've backtested your system, the optimal position sizing was definitely in excess of the position sizing enforced by IB? Under the example you just gave, what would your optimal position sizing be? Is there a determinate answer, or is just that more is better?

 

[Businessman]

Quote from whitemills:

So you are obviously trading an intraday, very short-term system. Does your system typically have you in all five markets at the same time?

I'm just trying to probe you on your position sizing... Its seems like you are basing your stops on your position size and the maximum number of contracts allowed by the margin rules.

Just to be clear - you are stating that when you've backtested your system, the optimal position sizing was definitely in excess of the position sizing enforced by IB? Under the example you just gave, what would your optimal position sizing be? Is there a determinate answer, or is just that more is better?

My system does not use fixed or optimal position sizing.
It is based on volatility, but it adapts faster than the exchange margins do. It has increased sizing now that the VIX and daily ranges have come down. But the exchanges have not reduced margins to reflect this. If IB still had intra day margin it would be ok.
Like i said i can still trade two or three market at full size, but if i get a signal in all of them at the same time then i have to handle that by cuting size below what i would like.
When the VIX or volatility increase again then the system will cut back size and it will be ok again.

 

[whitemills]

So at times, your system would have you all-in at times with a position size somewhere between the what is allowed by the current IB margins and 2x that level (50% intraday margin)?

I just hope your system is really accurate and you are damned confident in those inter-market correlations. Note that the implied leverage of the NQ contract at $4K margin was 6.3:1 today, and 12.6:1 at the margin at which you desire to trade. I think some of your other markets might even be a shade higher. That is a pretty big knife if a few swings hit you the wrong way.

I don't know the details of your system and I have no reason to believe that you aren't a capable trader, but I leave you with one parting thought: realize that your system seems to adjust very quickly based on observed short-term historical volatility. Have you calculated the correlation over a historical data set between historical observed volatility and volatility going forward?

Low observed volatility over the past decade in real estate prices, defaults, etc. led to the securitization markets letting people leverage mortgage investment by 50x to 100x. We all know how that ended. You should think about incorporating some stress-testing based on empirical spikes from past event risk, rather than a simple trailing observed volatilities.

Would you write earthquake insurance based on the past 3 years of data? On Wall Street, the answer is yes, since my bonus is annual

Happy trading

 

[Businessman]

Quote from whitemills:

So at times, your system would have you all-in at times with a position size somewhere between the what is allowed by the current IB margins and 2x that level (50% intraday margin)?

I just hope your system is really accurate and you are damned confident in those inter-market correlations. Note that the implied leverage of the NQ contract at $4K margin was 6.3:1 today, and 12.6:1 at the margin at which you desire to trade. I think some of your other markets might even be a shade higher. That is a pretty big knife if a few swings hit you the wrong way.

I don't know the details of your system and I have no reason to believe that you aren't a capable trader, but I leave you with one parting thought: realize that your system seems to adjust very quickly based on observed short-term historical volatility. Have you calculated the correlation over a historical data set between historical observed volatility and volatility going forward?

Low observed volatility over the past decade in real estate prices, defaults, etc. led to the securitization markets letting people leverage mortgage investment by 50x to 100x. We all know how that ended. You should think about incorporating some stress-testing based on empirical spikes from past event risk, rather than a simple trailing observed volatilities.

Would you write earthquake insurance based on the past 3 years of data? On Wall Street, the answer is yes, since my bonus is annual

Happy trading

I understand what you are saying.
But my real secret weapons are my 'stop loss' orders, i think it is Taleb who in one of his books says that the use of stop losses to manage risk on Wall Street is quite rare.
Now i realise that the market can gap through my stop losses. But it would have to in all the markets at the same time to really hurt me.
My biggest risk is if globex goes down while i am in three markets at the same time and reopens with huge losses in one or all and i wasnt able to hedge while it was down.
Although there is a chance of that happening i think the odds would be similar to me getting killed in the real world, by a terminal disease, car crash etc.

 

[whitemills]

Well... I think one of the underlying themes that you also get from everyone from Taleb to Buffett is the fundamental misunderstanding in modern financial theory of the difference between volatility and risk. It still makes me cringe when I read investment literature aimed at newbies that talks about 'beta' as a measure of how risky an asset is.

Just to confirm - you are trading a very fast system that usually has you in and out of the market within hours if not much sooner? How long have you traded this system in real life and/or how long have you tested it back over? I would expect the impact of commissions and slippage to be quite large.

I guess my worry for you stems from the fact that it seems that your system requires high amounts of leverage to generate the returns you seek, or else we wouldn't be having this discussion... if life was as simple as finding any method with positive expectancy and then just seeking the maximum possible leverage, trading would be a much easier endeavor.

For reference, i trade with an equity size of several hundred thousand and I have had the glorious experience in the past of flaming out an account because I was trading too large. I'm not into schadenfreude, so I'm just giving you my honest counsel. I'm just very skeptical that you can trade at the kind of leverage you are contemplating without experiencing 10% or larger daily drawdowns frequently, and that is only if your system is very good.

Two things:
(1) Even if your test show validity of a concept 99% percent of the time, that 1% would translate into 2-3 days per year that would fall outside of you bounds. That is clearly a higher frequency than you being hit by a car, and 99% is a damn good confidence level in any business.
(2) If you are trading at or near max leverage, it takes you a lot longer to recover from losses... i.e., if you are trading your limit of 40 contracts, then take a loss, all of a sudden you take a 10% loss, you only have enough equity left to trade 36 contracts, so you need a bigger move than what dropped you down just to get back to breakeven. This is a simplified explanation of the first step in the downward spiral that claims a lot of aggressive traders.