MK
1) Is there any possibility of a new classification category of systems - such as:
– Conservative Risk (leverage factor <1.5) - which includes unleveraged stock trading. These systems rarely or never exceed the leverage factor of their category (as in the example of the system "Conservative Forex, which was obviously not conservative)
– Moderate Risk (leverage factor 1.5 to < 3.0) - includes trading of things like ProFunds/Rydex Funds, which usually leverages up to 2.5)
– Average Risk (leverage factor 3.0 to < 5.0) - a reasonable leveraging for futures/forex trading
– Risky (leverage factor) (leverage factor 5.0 to < 10.0) -
– Extreme Risk (leverage factor >= 10) such as some forex trading or heaviliy leveraged futures/other trading, where someone uses the $500 margin to load up on eminis, etc. These are often all or nothing systems.
This can be a way for prudent system vendors to earn the respect of traders looking for a system that doesn’t constantly roll the dice while trying to roll up the equity curve (a la Flying Pink Pig). Leverage is an excellent measure of system/vendor maturity.
The word “Conservative/Moderate/etc.” can appear as a column in the Grid, and /or as a RIsk Measure under statistics or elsewhere on the page. It can also be a categorizer such as on the lefthand side such as Best Systems. You can put “System Risk: Conservative Moderate Risky” whatever. Up to you.
2) Perhaps you can even consider a new metric - "leverage factor"LF as
– per trade LF which indicates the total leverage for a trade, as % of available total trading capital and/or
– total LF (TLF) - which indicates total leverage for all current trades, as % of available total trading capital and/or
– average LF (ALF) which indicates average leverage for the life of the system.
Right now, it is difficult to determine how heavily leveraged a system is, and that is often the cause of system collapse, as in May 2006, Feb 2007, and the past couple of weeks, when the market tanks.
Probably the only constructive thing he has to say. I would support this as well. I have tried to calculate this as the percent change in the instrument divided by the percent change in the index/security and its impact on equity to indicate the extent of the leverage. Obviously this will depend on the underlying’s volatility and the movement of the instrument itself and what percentage change that is on the equity. I guarantee some leverage factors would exceed 100. Options are a tougher call because the instruments are inherently volatile, but we would be talking about the percent change in the instrument and how that affects the equity.
I had a calculation for futures discussed on WL, but had no answer to it. Something if, say, a futures contract changes a point, what is that on a percentage basis divided by the percentage change in, either the invested position only, or as the percentage change in the equity curve. Either way you do it, I believe most futures contracts are leveraged at least 20:1 by this measure.–
Beau
I understand the intent, but I don’t see a meaningful way to do it. Even in the stock arena, it isn’t straightforward. Suppose someone’s trading “long only” at account equity (totally invested but no use of margin). His factor would be 1. But if he’s a long/short trader, or a long/hedged with puts trader that’s fully invested, his number would also be 1, but in those hedged cases I’d argue he should have a number less than 1.
Going to Forex or futures, it is much harder to make a meaningful comparison.
Leverage factor is great!!
I think its a nice metric to show- if a vendor has great annualized annualized return and stats, but doen’t use leverage- its a sign of strength.
Also- I would also like on the closed trades that everyone sees to see what % of portfolio value that was traded…
So it doesn’t make sense to you to say, if the underlying moves this much percent, how much percent does that move equity? In the case of the long/hedged with puts trader, the puts would offset some of the volatility assuming the number of puts is perfectly matched. Meaning if he owns 100 shares, 1 put would cover his position exactly and both securities would move inversely with one another, thereby reducing leverage. I think it could be taken more like, a percent change in each security translates into what percentage change in the equity. That is, if you own 15 securites unhedged at 10% of equity, obviously the leverage factor would be 1.5 initially, but as the positions fluctuated up or down, the leverage factor would also vary.
What I was saying in WL was, if 1 point in NQ is worth $20, 1/1774=x%, the ratio of %changeequity/x% is the leverage factor. Depending on how many contracts you buy, you’ll get the leverage factor this way. You would use an options pricing model to guess the percent change in the option price and probably get a very close estimate just using the usual variables to price the option. For that, we have %changeinstrument=x%. %changeequity/x% would be the leverage factor in that case as well.
You would agree, though, that if someone invests 100% of his capital into a security exactly, and if the price drops $0.01 equity declines $0.01, and correspondingly 1% is 1% lost. I am not understanding why this approach does not work. I guess in the case of leveraged ETF’s the program would have to search for the underlying index to properly compute leverage. But I think the process would work for futures stocks and options.–
Beau
"But if he’s a long/short trader, or a long/hedged with puts "
That is true of many metrics - eg, spread traders. But this is a small minority of the systems here, and does not take away from the difficulty of evaluating systems from a leveraged (danger) piont of view
"Options are a tougher call "
True, but like the long/short objetion, the vast majority of C2 systems are Forex pairs, plain futures, plain stocks, etc. This is what the leveraging factor (LF) metric is trying to measure.
Right now, it is a crap shoot, and many vendors abuse leverage, and make it difficult for newer traders to avoid the catastrophic losses in markets such as the past few weeks. Six Sigma is not the only implosion.
"So it doesn’t make sense to you to say, if the underlying moves this much percent, how much does that move equity?"
No that’s perfectly clear. And for single instrument systems say a QID-, or @es-trader it’s pretty straightforward. But for mutliple position strategies, it’s much like the drawdown calculation. You have to measure the market moves and equity moves every second to measure cause and effect, because some are positively correlated, and some negatively correlated.
I think that complicated empirical risk calculations, like the ones suggested in this thread, are beyond what can be expected of C2. I suggest to simply use the IB margin requirements. I would expect that these reflect the involved risk reasonably well. The leverage factor would be defined as margin requirement / total account value. A value close to 1 means that the account is close to a margin call. Even if you disagree with IB’s risk assessment, this will be a fact that no one can ignore.
This will not capture the leverage involved in options and funds like QLD. On the other hand, ordinary stocks do also have different degrees of volatility, and I believe that trying to estimate this on a daily basis for all possible stocks and options is like opening a can of worms. If this was easy, IB would probably do it too.
Jules - I agree re the keep-it-simple approach. The main issue is Total Portfolio Leverage as risk in individual trades can be controlled by trading liquid instruments and using auto-stops (% balance at risk based).
I agree with Ross re the bang-and-bust type systems seen on C2, and that it is not appropriate for a conservative risk client to sign up for a high leverage system - even though C2 doesn’t give advice, it should give key information so folk can make informed decisions.
The interest shown in this thread, and an earlier one I started on the same subject (Dec 10, 2007) should tell Matthew this is a desired feature.
In that earlier thread you made an excellent suggestion:
"I suggest to compute leverage in exactly the same way as it is done in an IB account, in TWS > Account information > Available for trading > leverage. According to the TWS user guide, this is defined as
leverage = Gross Position Value/Net Liquidation
But, if I understand the TWS user’s guide, this does not inlcude margin requirements. That would be another measure of interest, something like
total margin requirement / account value."
I think both Leverage and Margin Requirement calcs should be straightforward and could be computed on a daily snapshot basis.
I’d like to see, for Leverage and Margin Requirement:
* highest this month
* % change on highest last month
* highest in system life
Surely this site can handle it on a day to day basis or from initial entry, right? I think it is as simple as just doing the calculation from the initial entry. I guess I went a little far suggesting a continuous update, which I do acknowledge is not feasible.
