Potential new financial system

[Godders]

Hi all, For a while now, I've been interested in financial spread betting, and have been researching it for the last few years. I think I've now come up with a system that I'm happy to trial, and if there are no objections, I'd like to trial on here. I do fully appreciate that this is more a sports betting forum, so if people aren't happy to have a financial system on here, then I'll happily trail it privately. Reason I'd like to make it public at this stage is to A) provide some motivation to actually properly do the research, and B) look for other people's opinions given I'm still learning my way through the markets! What am I proposing? This is primarily a trend-following system. I will identify trends on weekly charts, and look for "value" entry and exit points on daily charts. This means that positions will be held for a number of days, but they will be opened in the direction of long-term trends (i.e. a Buy position if the weekly chart is showing an uptrend, a Sell position if showing a downtrend). Positions will be opened as I would if were spread betting for real. Any equity listed on the FTSE 100 is a potential candidate for trading. How will I measure success? I have a couple of proposals here. I would like to provide some form of yield calculation for direct comparison to other systems on here. However, I'm unsure which way to go... I could use Muppet77's yield calculation from his Total Corner Spreads thread (i.e. profit / (money won + abs(money lost)) (credit Muppet77)). Or, since I will be placing a stop-loss on every position I open, I will effectively have a set amount of risk on any one trade. I could, potentially, measure yield as Total profit/total maximum risk - I think this calculation is more in line with the yield calculation for fixed odds systems on here. Does anyone have any opinions on which way would be best? Money-management First off, I would like to point out that I have absolutely no investment experience, and that this will be a PAPER-TRAIL ONLY! I do not advocate putting any real money on my selections. The starting bank will be £3,500 (as this would likely be my starting bank if I decided to do this for real tomorrow), and the maximum downside risk will be 2% of the bank. How is downside risk calculated? Each trade will have a stop-loss set when the trade is placed. This is an automatic instruction to close the position if a set price is hit. The idea being that if the trade moves against you, you will only lose the difference between the current price and the stop-loss price multiplied by the stake per point. As an example, lets say I want to open a position in a share that's trading at 500 points, and I will place the stop-loss at 450 points. In theory, the maximum I can lose on this trade is (500 - 450) = 50 points. I want my maximum risk to be 2% of £3,500 (or £70), so I will BUY 500, at (£70/50) = £1.40 per point (in spread betting on UK Equities, 1 point = 1 pence. I find it easier to think in terms of points so I don't get confused between pence and pounds!) Overnight Financing It is almost 100% likely that I will have to leave my positions open overnight. This means they are liable for an overnight financing charge levied by the spread betting firm. The calculation is usually the total value of the position at the close of the markets, multiplied by the Base interest rate (set by the Bank of England) + 2%, then divided by 365 (i.e. to give the daily finance charge). As an example, lets say that I've opened my position as above, and at close of trading, the share has risen to 510 points (the current bast rate is 0.5%). The total value of the position is £1.40 per point * 510 points = £714. The annual financing charge will be £714 * 2.5% = £17.85. This means that my overnight financing charge on this position will be £17.85/365 = £0.05. This calculation is done on a daily basis and on all open positions, so when calculating yield, I will use profit/loss after all charges have been taken. I'm currently backtesting this over the last few years, but so far it's looking promising. Assuming that there are no objections to this, I plan to start on 4th Jan, when the markets open after the new year.

[clay747]

Re: Potential new financial system Sounds interesting, I am keen to learn more about financial spread betting, so I will be following with interest. Do you plan on sharing the details of the system with us? or just proofing your bets? You may want to drop by muppets corner thread, we are discussing how best to calculate ROI for spread betting systems at the moment. Best of luck mate.

[v-zero]

Re: Potential new financial system The yield of a spread bet can easily be given once a position is closed, it is simply: (sell price/buy price) - 1. Hope that helps.

[clay747]

Re: Potential new financial system

The yield of a spread bet can easily be given once a position is closed, it is simply: (sell price/buy price) - 1. Hope that helps.

That works for buys, how about sells?

[Mulkis]

Re: Potential new financial system Very interesting indeed. I've been trading on financial markets for a couple of years already and find a lot of similarities between financials and sports-trading (especialy in risk management area). Will certainly follow this tread. As for now I would suggest to pay some attention not only to stop-loss orders, but to profit taking as well. I mean, the profit taking should also be as automatic as possible. Trend-following in general is probably the best way to play this game, but you need to have clear triggers for everything, starting from enter points to exits. Cheers!

[v-zero]

Re: Potential new financial system

That works for buys' date=' how about sells?[/quote'] Position yield is only measured on closed positions. Closed positions always have both a buy and sell price, so the equation is valid for positions that were initially sells and those that were buys.

[Godders]

Re: Potential new financial system Cheers for the reply folks - looks like I'll be starting this in the new year! I plan to give daily updates on new/existing positions, and weekly updates on trending shares (i.e. potential candidates for new positions in the upcoming weeks). I'll also do my best to update performance figures on a weekly basis. Mulkis - I agree that things should be as mechanical as possible, at least initially! Although I also think that part of this exercise is to gain confidence in dealing with the markets, and be able to form an opinion on what is happening to give an extra 'edge' when dealing with pure systematic traders.

The yield of a spread bet can easily be given once a position is closed, it is simply: (sell price/buy price) - 1. Hope that helps.

V-Zero, please can you let me know where you got this from? Looking at the equation, I think it's measuring profit on a trade as opposed to the yield? In my opinion, yield should measure the amount of profit as a percentage of the amount risked on any particular trade. A yield of 5% for example suggests that for every £100 risked, you could expect £5 profit on average. Correct me if I'm wrong though! Thanks very much for the input V-Zero, but I don't think that, in it's current form, this equation is doing what I want it to do, in that there's nothing relating the profit made to the amount that was risked.

[v-zero]

Re: Potential new financial system

Cheers for the reply folks - looks like I'll be starting this in the new year! I plan to give daily updates on new/existing positions, and weekly updates on trending shares (i.e. potential candidates for new positions in the upcoming weeks). I'll also do my best to update performance figures on a weekly basis. Mulkis - I agree that things should be as mechanical as possible, at least initially! Although I also think that part of this exercise is to gain confidence in dealing with the markets, and be able to form an opinion on what is happening to give an extra 'edge' when dealing with pure systematic traders. V-Zero, please can you let me know where you got this from? Looking at the equation, I think it's measuring profit on a trade as opposed to the yield? In my opinion, yield should measure the amount of profit as a percentage of the amount risked on any particular trade. A yield of 5% for example suggests that for every £100 risked, you could expect £5 profit on average. Correct me if I'm wrong though! Thanks very much for the input V-Zero, but I don't think that, in it's current form, this equation is doing what I want it to do, in that there's nothing relating the profit made to the amount that was risked.

It's simply the only sensible way of calculating yield on a spread bet whose minimum points value is zero. When you 'buy' a position you essentially tie your stake per-point to a point value. The minimum value that your points position could fall to is zero, therefore the amount you stand to lose, your stake if you will, is simply the points value of your buy position, multiplied by your stake per point. Now, spread-bets on financials are ongoing, so they have no distinct settling points. As such, all buy orders are eventually closed with sell orders, and vice versa. Look at in the rather simple case of purchasing and then selling an apple. Say you buy that apple for £1, then you want to sell it. If nobody is willing to buy it, your apple is worth £0 and you have lost your entire investment. If, however, you are able to sell on the apple for £1.10, then your return on a £1 investment is 10%, i.e. ((1.1/1) - 1)*100... Likewise, if you have an apple, and sell it for £1.50, then you are able to buy it back for £1, then your sell price over buy price is 1.5, and so your return on investment is 50%. That makes sense, since in gaining 50p you have not lost the apple, the apple you valued at 100p, so you can say you've gained half an apple on your original single apple investment, in terms of sheer intrinsic value, which is all the spread betting points values are, they are a representation of the present 'value' of a commodity, in your case a financial commodity, which you can then make predictions on and either buy the commodity, believing it to be likely to grow in 'value' over time, or you can sell it, believing that it will lose value over time. I hope this makes some sense. EDIT: I thought I would just point out that there is absolutely no difference between spread betting and betting in play in betfair, the two are entirely equivalent, the only differences being the way positions are represented on the screen, and the fact that betfair markets eventually settle - they are not ongoing. The equivalent points value of a decimal odds bet is: points value = 1/(decimal odds - 1). Obviously this equation can be used both ways. e.g. points value of a 2.0 decimal odds bets is 1. The points value of a bet at 5 is 1/4 = 0.25. The points value of a bet at 1.01 is 1/0.01 = 100 etc etc... Decimal odds equivalent of a 100 point spread bet is either a back or lay at 1.01 depending on whether you back (buy) or lay (sell).

[Godders]

Re: Potential new financial system Thanks for the explanation V-Zero. However, I'll give a couple of examples to hopefully explain my argument a bit better: Example 1 - BUY a share trading at 100 points. Stop loss will be placed at 50 points, and maximum risk will be £100. Stake per point will therefore be £100/(100-50) = £2 per point. I then close the position at 110 points. Your calculation shows a yield of 10% on the trade. Example 2 - BUY a share trading at 100 points. Stop loss will be placed at 75 points, and the maximum risk will be £100. Stake per point will therefore be £100/(100-75) = £4 per point. I then close the position at 110 points. Your calculation shows a yield of 10% on the trade. Hopefully, you can see the introduction of the stop loss on these trades has introduced a big difference between them - in example 1, the profit on the trade is 10 points * £2 per point = £20. Example 2 is showing a profit of £40. Since the downside risk is the same on both trades (£100), and the profit is different, I would expect the yield to differ between the two trades too - this is what your calculation of yield doesn't show. My (simplistic) way of measuring yield outlined in my first post is profit/maximum downside risk. In these examples the yield would be: Example 1 - (£20/£100)*100 = 20% Example 2 - (£40/£100)*100 = 40% Total yield after these 2 trades would be (£60/£200)*100 = 30% I've just taken Clay's advice and dropped by Muppet's corner thread to look over the arguments on there. It looks like that thread is at a more advanced level of discussion than this is at the moment, so I suggest discussing any further subtleties of yield calculation on spread bets over there for the time being! Given that I've still got over a week before I'll be starting this system, I'll see what Muppet's thread comes up with in terms of a yield calculation before confirming exactly how I'll be measuring my success. Ideally, I'll be doing it on the same terms as Muppet to provide a direct comparison.

[v-zero]

Re: Potential new financial system

Thanks for the explanation V-Zero. However, I'll give a couple of examples to hopefully explain my argument a bit better: Example 1 - BUY a share trading at 100 points. Stop loss will be placed at 50 points, and maximum risk will be £100. Stake per point will therefore be £100/(100-50) = £2 per point. I then close the position at 110 points. Your calculation shows a yield of 10% on the trade. Example 2 - BUY a share trading at 100 points. Stop loss will be placed at 75 points, and the maximum risk will be £100. Stake per point will therefore be £100/(100-75) = £4 per point. I then close the position at 110 points. Your calculation shows a yield of 10% on the trade. Hopefully, you can see the introduction of the stop loss on these trades has introduced a big difference between them - in example 1, the profit on the trade is 10 points * £2 per point = £20. Example 2 is showing a profit of £40. Since the downside risk is the same on both trades (£100), and the profit is different, I would expect the yield to differ between the two trades too - this is what your calculation of yield doesn't show. My (simplistic) way of measuring yield outlined in my first post is profit/maximum downside risk. In these examples the yield would be: Example 1 - (£20/£100)*100 = 20% Example 2 - (£40/£100)*100 = 40% Total yield after these 2 trades would be (£60/£200)*100 = 30% I've just taken Clay's advice and dropped by Muppet's corner thread to look over the arguments on there. It looks like that thread is at a more advanced level of discussion than this is at the moment, so I suggest discussing any further subtleties of yield calculation on spread bets over there for the time being! Given that I've still got over a week before I'll be starting this system, I'll see what Muppet's thread comes up with in terms of a yield calculation before confirming exactly how I'll be measuring my success. Ideally, I'll be doing it on the same terms as Muppet to provide a direct comparison.

I thought I had given enough information to imply how calculations would be in in lieu of a stop loss, but I suppose not. A stop loss has the effect, of course, of changing the bet. It rebases the entire index. So, if the index has a lower end of 0, but you buy at 100 and will always leave at 75, then you have rebased the index such that a 100 is a 25, a 130 is a 55, a 200 is a 125 etc... You have to use rebased index prices to correctly calculate yield when a stop-loss is in place. Take your examples: Buy at 100 with a stop at 50, selling at 110. Rebasing gives a buy at 50, sell at 60. Hence your yield is (60/50 - 1) = 20% Buy at 100 with a stop at 75, selling at 110. Rebasing gives a buy at 25, sell at 35. Hence your yield is (35/25 - 1) = 40% The calculations you have made are correct, they are simply needlessly complicated by involving your stake-per-point value. If this is of use to muppet's thread then hopefully it will be seen. This is, I can assure you, the only correct way to measure yield equivalently to how yield is measured on binary bets.

[clay747]

Re: Potential new financial system

Position yield is only measured on closed positions. Closed positions always have both a buy and sell price' date=' so the equation is valid for positions that were initially sells and those that were buys.[/quote'] buying and then selling is not the same as selling and then buying. If you look at the profit graphs of buy/sell and compare it to a sell/buy you will see that when you buy first you have a fixed downside, prices can not fall below zero, your profits are unbound, prices can keep rising with no max price. When you sell first your profits are now fixed, prices can not fall bellow zero, but your down side is now unbound, the amount you can lose will keep rising as the price increases with no max price to stop at. If we look at the formula (sell price / buy price) - 1 when we buy first, the buy price is fixed and the sell price variable. as the sell price approaches zero the fraction sell price/ buy price will approach zero. so the maximum loss the formula will allow is (0) - 1 = -1 that is -100%. Our buy price is our investment, the amount at risk, we can never lose more than 100% of our investment. as the sell price approaches infinity the fraction sell price / buy price approaches infinity. (inf) -1 = inf. our profits unbound. however, when we sell first, the sell price is fixed and the buy price is variable. as the buy price approaches zero the fraction sell price/ buy price will approach infinity, (inf) - 1 = inf, this make no sense, our fixed profit has produced and ROI of infinity? as the buy price approaches infinity the fraction sell price / buy price will approach zero, (0) - 1 = -1. A max ROI of -100%. Our unbound loss has a max ROI of -100%? this only makes sense if the amount at risk is our account(our net betting wealth) once our losses pass that amount the bookie/broker will put a limit on our losses; and demand payment! applying the formula the way you suggest to sell bets, means what ever the close price was, that's the amount you risked. without a predefined stop loss the close price is just random number. So what meaning does your ROI have when your at risk is a random number?

[v-zero]

Re: Potential new financial system

buying and then selling is not the same as selling and then buying. If you look at the profit graphs of buy/sell and compare it to a sell/buy you will see that when you buy first you have a fixed downside, prices can not fall below zero, your profits are unbound, prices can keep rising with no max price. When you sell first your profits are now fixed, prices can not fall bellow zero, but your down side is now unbound, the amount you can lose will keep rising as the price increases with no max price to stop at. If we look at the formula (sell price / buy price) - 1 when we buy first, the buy price is fixed and the sell price variable. as the sell price approaches zero the fraction sell price/ buy price will approach zero. so the maximum loss the formula will allow is (0) - 1 = -1 that is -100%. Our buy price is our investment, the amount at risk, we can never lose more than 100% of our investment. as the sell price approaches infinity the fraction sell price / buy price approaches infinity. (inf) -1 = inf. our profits unbound. however, when we sell first, the sell price is fixed and the buy price is variable. as the buy price approaches zero the fraction sell price/ buy price will approach infinity, (inf) - 1 = inf, this make no sense, our fixed profit has produced and ROI of infinity? as the buy price approaches infinity the fraction sell price / buy price will approach zero, (0) - 1 = -1. A max ROI of -100%. Our unbound loss has a max ROI of -100%? this only makes sense if the amount at risk is our account(our net betting wealth) once our losses pass that amount the bookie/broker will put a limit on our losses; and demand payment! applying the formula the way you suggest to sell bets, means what ever the close price was, that's the amount you risked. without a predefined stop loss the close price is just random number. So what meaning does your ROI have when your at risk is a random number?

You cannot consider ROI on open bets. ROI is a result of a closed bet, and hence it doesn't matter if you bought first or not. Everything you pointed out in your post is well defined, you've not said anything that places doubt on the accuracy of the yield calculation. Selling first does indeed give you no theoretical loss limit, and buying first does indeed give you one, but a closed position is a buy and a sell, and since ROI is measured once positions are closed, it does not matter which came first. The fact that selling first can theoretically give an infinite loss is pointed out by spread betting firms, who are obliged to state that spread bets can face 'unlimited losses greater than the balance of your account'.

[clay747]

Re: Potential new financial system

You cannot consider ROI on open bets. ROI is a result of a closed bet, and hence it doesn't matter if you bought first or not. Everything you pointed out in your post is well defined, you've not said anything that places doubt on the accuracy of the yield calculation. Selling first does indeed give you no theoretical loss limit, and buying first does indeed give you one, but a closed position is a buy and a sell, and since ROI is measured once positions are closed, it does not matter which came first. The fact that selling first can theoretically give an infinite loss is pointed out by spread betting firms, who are obliged to state that spread bets can face 'unlimited losses greater than the balance of your account'.

the closing of a bet defines your profit. the opening of a bet defines your investment. when you buy you know how much your are risking, when you sell you should know how much you are risking, it does not depend on where you close the bet. if we sell corners at 10 and the game finishes with 1 corner you say the ROI is (10/1)-1 = 9 = %900 if we buy corners at 10 and the game finishes with 19 corners you say the ROI is (19/10)-1 = 0.9 = %90

[muppet77]

Re: Potential new financial system great discussion. is there any consensus on how best to compare spread systems? mine are always sells with a stop loss of ten corners. the sell value is usually around 11.

[v-zero]

Re: Potential new financial system

great discussion. is there any consensus on how best to compare spread systems? mine are always sells with a stop loss of ten corners. the sell value is usually around 11.

Time-based ROI is a pretty good measure... EDIT: For sell first bets with a stop price, the ROI is given by: ROI = (buy price - stop price)/(sell price - stop price) - 1 This is because of the 'direction' of the bet.

[muppet77]

Re: Potential new financial system so how would that work for multiple bets?

[v-zero]

Re: Potential new financial system

so how would that work for multiple bets?

Yield of multiple bets = sum of individual bet yields/number of bets.

[muppet77]

Re: Potential new financial system ok - i get +18% from 323 bets???? is that good?

[v-zero]

Re: Potential new financial system

ok - i get +18% from 323 bets???? is that good?

Well, it's a fantastic yield, if that's what you're wondering. :)

[muppet77]

Re: Potential new financial system brilliant, so now its directly comparable to fixed odd systems?

[v-zero]

Re: Potential new financial system

brilliant' date=' so now its directly comparable to fixed odd systems?[/quote'] Yes. If you wanted you could even convert each of your bets to a fixed odd equivalent form after the position is closed.

[muppet77]

Re: Potential new financial system in fact i forgot to put in the stop loss part to the eqn and it stands at +24%

[muppet77]

Re: Potential new financial system v-zero, what if i sell at 11 corners with a stop loss of 10. there was 1 corner in the game. ((11-1) / (1-1)) -1 = ?????

[v-zero]

Re: Potential new financial system

v-zero, what if i sell at 11 corners with a stop loss of 10. there was 1 corner in the game. ((11-1) / (1-1)) -1 = ?????

I think you've misunderstood my equation. Your stop price on a sell bet is your sell price plus your maximum loss. In this case your stop price is 21.

[muppet77]

Re: Potential new financial system so it would be (11-21) / (1-21) -1?

[clay747]

Re: Potential new financial system

so it would be (11-21) / (1-21) -1?

ROI = (buy price - stop price)/(sell price - stop price) - 1 so its (1-21)/(11-21) - 1 = 1 = %100 which agrees with profit / stoploss = (11-1)/10 = 1 = %100 however if we use the original formula we get, (sell price/buy price) - 1 = (11/1) - 1 = 10 = %1000 :eek

[muppet77]

Re: Potential new financial system ah, ok , the "stop price" is technically the "stop win" price rather than the "stop loss" on a sell bet. also the "buy" in the formula is actually the price taken as a "sell" by the punter. all clear now i think. i will rejig my formual and and see what i get out...

[muppet77]

Re: Potential new financial system rewriting the formula into idiot language (for my benefit) ((A - B) / (C - B)) -1 where A = total corners in match B = price taken as a "sell" from bookie + stop loss of 10 C = price taken as a "sell" from bookie how do you then take into account that some games will have a £2 stake / corner and some £3 / corner?

[clay747]

Re: Potential new financial system

rewriting the formula into idiot language (for my benefit) ((A - B) / (C - B)) -1 where A = total corners in match B = price taken as a "sell" from bookie + stop loss of 10 C = price taken as a "sell" from bookie how do you then take into account that some games will have a £2 stake / corner and some £3 / corner?

((A*stake - B*stake) / (C*stake-B*stake))-1 but that can be rewritten as (stake*(A-B)/stake*(C-B))-1 which simplifies to ((A-B)/(C-B))-1 that is, your corner unit roi will equal your £ roi whatever stakes you use.

[muppet77]

Re: Potential new financial system thanks clay. I get an answer of 5% :~(