Trading Mechanical Systems using Leverage

[Chorlton]

Hi All,

Not sure where to post this thread but hopefully this location is ok.

Has anyone developed any Mechanical Systems which were planned to be traded with leverage using either CFD's or via Spreadbetting??

If so, can anyone offer any help / guidence as to how one would go about testing a strategy which relies on leverage as opposed to one that doesn't? What differences should be considered, etc, etc.

Any help much appreciated,

Regards,

Chorlton

 

[0007]

Hi All,

Not sure where to post this thread but hopefully this location is ok.

Has anyone developed any Mechanical Systems which were planned to be traded with leverage using either CFD's or via Spreadbetting??

If so, can anyone offer any help / guidence as to how one would go about testing a strategy which relies on leverage as opposed to one that doesn't? What differences should be considered, etc, etc.

Any help much appreciated,

Regards,

Chorlton

Not as such, but.......

You can make your leverage proportional to the expectancy of your system. Eg high expectancy, you don't mind drawdowns - go for it. You can also factor in the drawdoen you're willing to accept. Is that what you mean? If I'm stating the obvious - apologies. Spreadbetting is leveraged and I factor my stake according to the parameters I quoted - though I do the calculation "on the back of a fag packet" & then record it with the trade.

I do believe some funds have got quite seriously caught out thru leverage though!

 

[Windowsill]

Hi All,

Not sure where to post this thread but hopefully this location is ok.

Has anyone developed any Mechanical Systems which were planned to be traded with leverage using either CFD's or via Spreadbetting??

If so, can anyone offer any help / guidence as to how one would go about testing a strategy which relies on leverage as opposed to one that doesn't? What differences should be considered, etc, etc.

Any help much appreciated,

Regards,

Chorlton

If spreadbetting is leveraged, which it is! then any system where spreadbetting is the betting vehicle has to be leveraged does it not?

IMO drawdown is drawdown. If you have x amount in your ac........simply work out how disasterous your system needs to be to ...using fixed or percentage bets....I use XL.

I apologise now if i have the wrong end of several logs.

 

[Chorlton]

Not as such, but.......

You can make your leverage proportional to the expectancy of your system. Eg high expectancy, you don't mind drawdowns - go for it. You can also factor in the drawdoen you're willing to accept. Is that what you mean? If I'm stating the obvious - apologies. Spreadbetting is leveraged and I factor my stake according to the parameters I quoted - though I do the calculation "on the back of a fag packet" & then record it with the trade.

I do believe some funds have got quite seriously caught out thru leverage though!

Hi 0007,

Thanks for your comments. The system I intend to use with leverage is posted here: http://www.trade2win.com/boards/mec...ng/28661-calling-all-system-developers-4.html (Post #36)

The system posted above was originally designed to be traded without leverage, however looking at its MaxDD, etc I believe there is now scope to use leverage to improve returns without affecting overall DD too much.

The reasoning for my original question was that I believe that one may have to consider additional / different considerations when designing a system which will be traded using leverage.

As an example, the original system uses a portfolio limit of 100% (Fully invested) whereas when trading with leverage this figure should maybe be reduced to cover possible Margin calls etc? Would you agree?

Finally, as the system was originally designed to trade stocks (without leverage) with a starting capital of £50,000, as I now intend to trade with approx 10-20% leverage (using Spreadbetting) does this mean I would only need 10-20% of my original start-up capital to fund my Spreadbet account?

Regards,

Chorlton

 

[Chorlton]

If spreadbetting is leveraged, which it is! then any system where spreadbetting is the betting vehicle has to be leveraged does it not?

IMO drawdown is drawdown. If you have x amount in your ac........simply work out how disasterous your system needs to be to ...using fixed or percentage bets....I use XL.

I apologise now if i have the wrong end of several logs.

Hi Windowsill,

Thanks for the reply.

Apologies in advance but could you expand on your comments, especially those in BOLD.

Cheers,

Chorlton

 

[counter_violent]

One of the biggest considerations currently would be volatility...as everything is winding up we ought to be sizing down.

 

[Windowsill]

Hi Windowsill,

Thanks for the reply.

Apologies in advance but could you expand on your comments, especially those in BOLD.

Cheers,

Chorlton

I'm still not sure I understand what you're asking.

To me its quite clear though, if your borrowing funds to trade with, you need to do a few calculations regards costs of borrowing, and incorporate them into your model, and ensure the profit margin is good enough to make it worthwhile.

The other thing I alluded to was the "risk of ruin", the calculation of which is somewhere on here, and again has to include the cost of funds.

Quite clearly you need to cover the costs, before you make a profit. The system does not look that robust to me,(not that i have a particularly good handle on what i'm looking at) ie: what if say five per cent of trades went against you instead of for you?

 

[0007]

Hi 0007,

Thanks for your comments. The system I intend to use with leverage is posted here: http://www.trade2win.com/boards/mec...ng/28661-calling-all-system-developers-4.html (Post #36)

The system posted above was originally designed to be traded without leverage, however looking at its MaxDD, etc I believe there is now scope to use leverage to improve returns without affecting overall DD too much.

The reasoning for my original question was that I believe that one may have to consider additional / different considerations when designing a system which will be traded using leverage.

As an example, the original system uses a portfolio limit of 100% (Fully invested) whereas when trading with leverage this figure should maybe be reduced to cover possible Margin calls etc? Would you agree?

Finally, as the system was originally designed to trade stocks (without leverage) with a starting capital of £50,000, as I now intend to trade with approx 10-20% leverage (using Spreadbetting) does this mean I would only need 10-20% of my original start-up capital to fund my Spreadbet account?

Regards,

Chorlton

Hi Chorlton,

I'm not a system builder so not qualified to comment on the merits of your system. General opinion seems to be that backtesting doesn't guarantee success but I think it might weed out a poor system. I'm not really sure how much leverage has to do with it as such - risk is risk and you can choose your level and put it into effect with varying levels of leverage. I think a point to remember is that over and above your chosen arithmetic value of risk, leverage in itself is an increased risk. Without leverage you know your absolute risk, otherwise the sky can be the limit - especially if you lack discipline. Also, most systems are complex and it's sometimes difficult to ascertain the effect of a changed variable. Yes there are backtesting facilities etc but if they were that good, lots of people would be using foolproof systems (maybe they are and just keep quiet?). But we hear of corporate and individual failures so what value can you put on backtesting which cannot test the future?

I would consider why and how much you want to use leverage: short of funds? want bigger returns? it's what "they" all do? - and then I would keep it to the minimum commensurate with desired risk profile. I personally only use leverage because that's the way SB works; but I don't risk any more than I would buying ftse stocks in a conventional way. For smaller amounts and frequent trading SB does save some cost (depends on your broker etc) but the big advantage for me is being able to go short in a convenient way.

Although I have a pretty reliable system (ie good expectancy) I trade very cautiously and don't over-leverage. I could make more dosh if I did, but could probably have gone bust by now. All depends on your outlook - if i want more money i just trade more, but i don't change my risk level or my system. A non-qualifying trade is a non-qualifying trade regardless of chosen risk or leverage.

In view of current markets (as counter-violent implies: volatility will be on the rampage) I think I'd tread very carefully. When i have tried new ideas / systems I first start paper trading then go real with very small stakes and then gradually wind up to a realistic (ie for me) level. I Always back off at any stage where it goes wrong and reconsider etc.

The master trader can probably make money with any half-decent system. Once I realised that it's discipline, objective assessment and risk management that are vital, I was heading down the right road.

Again, apologies if i've just stated the bl**ding obvious but that's been my experience and it might just benefit somebody.

 

[Chorlton]

I personally only use leverage because that's the way SB works; but I don't risk any more than I would buying ftse stocks in a conventional way. For smaller amounts and frequent trading SB does save some cost (depends on your broker etc) but the big advantage for me is being able to go short in a convenient way.

Hi 0007,

This is basically the reason why I am now looking at using Leverage. I do have the funds available to trade without leverage but prefer to use SB for specific reasons.

Regards,

Chorlton

 

[tss42]

A margined account simply lets you have more open positions than a cash account. That's all.

If you ever trade a fixed and limited number of markets and are never going to be anywhere near 100% invested, then it doesn't make a difference whether you have a cash or margin account. If however you strategy can generate a large number of trades in theory at any time, and you are likely to use nearly all your account equity on margin, then it has a direct impact on your ROE.

 

[nobel_1101]

Chorlton,

Personally I don't think it makes any difference to your backtesting whether you use a leveraged or unleveraged trading vehicle.
If you measure your backtesting results in terms of initial risk (R), you can set how much R is worth in terms of initial capital. E.g. R = 1.5% of total capital

The reason it doesn't matter whether its leveraged or not is because you only invest an amount which makes R = 1.5% (or whatever %age you're comfortable with)
E.g. say my system gave me a long signal to trade Barclays, entering at 450p with a stop loss at 420p. My unit risk is 30 points. Lets say I had £10,000 total capital.
Because I want to risk 1.5% on the trade I want my total initial risk to be £150.

If I want to spreadbet, that means I need to invest at £5 per point (i.e. £150 / 30 points = £5 per point).
If I want to buy shares directly, I need to buy 500 shares (i.e. £150 / £0.30 per share = 500 shares).

Whatever vehicle I use I'm still risking £150 or 1.5% of total capital.

Regards

Craig

 

[Hoggums]

Chorlton,

Personally I don't think it makes any difference to your backtesting whether you use a leveraged or unleveraged trading vehicle.
If you measure your backtesting results in terms of initial risk (R), you can set how much R is worth in terms of initial capital. E.g. R = 1.5% of total capital

The reason it doesn't matter whether its leveraged or not is because you only invest an amount which makes R = 1.5% (or whatever %age you're comfortable with)
E.g. say my system gave me a long signal to trade Barclays, entering at 450p with a stop loss at 420p. My unit risk is 30 points. Lets say I had £10,000 total capital.
Because I want to risk 1.5% on the trade I want my total initial risk to be £150.

If I want to spreadbet, that means I need to invest at £5 per point (i.e. £150 / 30 points = £5 per point).
If I want to buy shares directly, I need to buy 500 shares (i.e. £150 / £0.30 per share = 500 shares).

Whatever vehicle I use I'm still risking £150 or 1.5% of total capital.

Regards

Craig

Craig you may be risking £150 - but with Sb'ing you are commiting far less money to this trade. To Sb £5 per point on Barclays only requires £106 in your account at IG. To Buy 500 shares of barclays to give you the same profit costs over £2100.

So in theory you could take that remaining £1950 and enter a few more positions.

But what you have to consider is let's say you take out another 4 positions and the market falls - you'll lose the £150 on your shares - but with Sb'ing you'll lose £150 x 5 because of those extra positions you took. A 40% percent drawdown!

Also - it'll cost you everyday to hold those positions, You'll be charged £2100x5.25%pa on your barclays position (about 30p a day) for the money you've effectively borrowed in order to own these shares.

Let me put that in perspective as this is something that is often overlooked - if you invested £106 to buy those barclays shares - then you'd pay £2.10 every week to keep hold of them, yep - over 2% charges every week on your initial capital of £106.

 

[the blades]

Let me put that in perspective as this is something that is often overlooked - if you invested £106 to buy those barclays shares - then you'd pay £2.10 every week to keep hold of them, yep - over 2% charges every week on your initial capital of £106.

it's a fair point for anyone considering loading up on more stocks with the money they've "saved" through leverage. But an alternative is only exposing yourself to the same amount you would have done if you'd bought outright. Then stick the cash you save on the deal in the bank (offset a mortgage or whatever) and you're receiving more that the 5.25% your paying.

It's a great deal as far as I'm concerned

UTB

 

[nobel_1101]

Let me put that in perspective as this is something that is often overlooked - if you invested £106 to buy those barclays shares - then you'd pay £2.10 every week to keep hold of them, yep - over 2% charges every week on your initial capital of £106.

I take your point about the borrowing charges. They need to be factored into the backtesting. 2% of margined money looks high but put another way that only 0.021% of total capital. Considering you're risking 1.5% of total capital on the trade, and assuming its short/medium term position its not too much of a big deal.

You could take additional non-leveraged trades in that example as well but with a ceiling of £10,000 invested.

 

[Chorlton]

it's a fair point for anyone considering loading up on more stocks with the money they've "saved" through leverage. But an alternative is only exposing yourself to the same amount you would have done if you'd bought outright. Then stick the cash you save on the deal in the bank (offset a mortgage or whatever) and you're receiving more that the 5.25% your paying.

It's a great deal as far as I'm concerned

UTB

I agree, and this was one of the differences I see when developing a system to trade on leverage. Without leverage one can be "fully" invested but with SB'ing (as an example) it makes sense to keep some capital in cash, for the reasons already discussed. I see this as one of the DIFFERENCES that should be considered when deciding to use leverage.

On a slightly different note, when it comes to backtesting a strategy which will be traded using leverage (using SB'ing as an example), how do people deal with Transactions Costs / Commissions in backtesting???

As we know SB'ing is free of "fixed" commissions because instead the commissions are derived from increasing the spread of a particular stock.

However, depending on ones position sizing rules (and the difference in buy price & stop loss) the amount of points to be risked will vary at each trade.

As a result, in backtesting what value would you use for commissions??

Thanks in advance,

Chorlton

 

[Chorlton]

Also - it'll cost you everyday to hold those positions, You'll be charged £2100x5.25%pa on your barclays position (about 30p a day) for the money you've effectively borrowed in order to own these shares.
.

Hi Hoggums,

Can you expand on this please?

Excuse my ignorance, but I didn't realise that one would be charged interest in keeping a SBet open, or have I missed something here

Thanks.....

 

[nine]

Its just like margin Chorlton, if you have a position open with a leverage of 50:1 and your entire account invested you have a loan of 49 50ths of the position and are charged interest on the loan. Your spreadbetter will have a page somewhere that explains how they handle long and short positions.

 

[nobel_1101]

However, depending on ones position sizing rules (and the difference in buy price & stop loss) the amount of points to be risked will vary at each trade.

As a result, in backtesting what value would you use for commissions??

Hi Chorlton

That's a good question and also the spread offered by the Spreadbetting company varies by how liquid each stock or index is, which further complicates things.
I test using a pessimistic 1/3 of initial risk in costs and slippage across all trades but I guess it depends on how tight your stops are.
You could assume a fixed point (or fixed percentage) spread, then calculate the cost in terms of initial risk on a trade by trade basis.

Regards
Craig

 

[Hoggums]

Hi Hoggums,

Can you expand on this please?

Excuse my ignorance, but I didn't realise that one would be charged interest in keeping a SBet open, or have I missed something here

Thanks.....

Depends on how you're doing it - if you buy a future (SB or otherwise) then the price is higher than the spot by the equvalent you'd pay in interest and any divis you'd receieve - this difference slowly drifts down to the spot price at expiry.

If you buy the daily rolling then depending on the SB they will either deduct the interest from your account (e.g. cap spreads) or IG roll-over the daily opening at a slightly higher price than the close (or they close slightly lower than the actual - I don't know which) - thereby you miss out on a pip or two which they take as interest payment.

It works both ways though - if you short a share/indices then you will recieve interest payments.

This is the case whether you trade indices, FX, shares, bonds or any commodities.

Generally speaking it is also more expensive to hold daily rolling long term than the futures. CS charges an extra 2.5% for daily rolling over the base rate.

 

[Chorlton]

I test using a pessimistic 1/3 of initial risk in costs and slippage across all trades but I guess it depends on how tight your stops are.

Regards
Craig

Hi Craig,

How do you go about "testing"? Do you use some additional software or using excel for example?

Thanks....