27/03/2009 Positions taken

Sorry for posting this now ... and already day has ended at London. Shortly I will also post another post with report on how our trades have fared.

Our picks of the day were:

Barclays Call 2009/09 2.50 R268 1.50 +2.09 4.8 +33.95
Barclays Call 2009/09 2.00 R267 1.50 +2.09 3.1 +24.32
Lloyds Call 2009/09 1.00 R321 0.76 +1.68 4.8 +18.10
HSBC Call 2009/09 5.23 R272 4.10 +3.33 2.3 +10.66
Vodafone Put 2009/09 1.20 R200 1.18 -2.75 2.3 +1.49
Tesco Put 2009/09 4.00 R337 3.30 -2.81 1.1 +1.11
BP Call 2009/09 5.00 R203 4.80 +4.46 2.2 +0.03

Their offer prices were picked up from saved files as I described in earlier ilog. Below I am giving offer prices along with source file and share name:

2009-03-27-10-28-38 R268 0.22
2009-03-27-10-33-27 R267 0.34
2009-03-27-10-38-22 R321 0.21
2009-03-27-10-44-06 R272 0.40
2009-03-27-10-48-22 R200 0.23
2009-03-27-10-59-21 R337 0.94
2009-03-27-11-06-21 R203 0.47

We had cash 4814.89 GBP. Today there are 7 prospects. Invest 687.84 on each prospect. Deduct 12.5 for commission: 675.34.

Shares now we own in paper trading account with share name, how many bought, price, total cost:

R268 3069 0.22 687.68
R267 1986 0.34 687.74
R321 3215 0.21 687.65
R272 1688 0.40 687.7
R200 2936 0.23 687.78
R337 718 0.94 687.42
R203 1436 0.47 687.42
Cash 1.5

-----

The problem of size

I think so both RBS and my broker are not doing it right by not having maximum size correctly being shown. Sometimes maximum is less than ordered shares still a quote is provided. In real life this is going to be problem. If after all the trouble of picking a share to invest if you are hit by order failure then how it is going to work? In real life I propose that when order fails with the amount you requested do not try to reduce amount. That is because you may be deciding on size based on other algorithms. Go back and recalculate smaller size's profitability again. With smaller size commissions are going to be bigger and profitability smaller. So consider that also. If commission cost is only consideration, make sure that at least bid-offer spread loss can be erased by profit indicator you are calculating. For example my Barclays prospect (Call, 2009/09, 2.50, R268) has profit potential of 33.95%. Frist we need a forecast model to predict from this number average actual gain that will happen. Supposing it predicts 3.4%. Now (say) we want to set target of 2% nett. Then what is the minimum size required to generate 2% gain if actually 3.4% is realised? Calculate that size. Remember that actual average will be based on (hopefully) many trades which realise target and (hopefully) few trades which we liquidate at just before close of day. Not necessarily setting higher target translates into higher average gain. Higher the target less likely share reaches that high and we may end up liquidating at lower or negative gain. When all these trades are accounted average will be less than target. If your system generates p% average nett gains, then asymptotically optimal target is p% and actual gain is higher than p% because for poor performers we just wait till next day and let price process generate p% gain eventually (that being the average). But a percent of trades close at lower gains in our system. Therefore, optimal target is less than p%. May be p% minus half of variance?

Another thing to note is equi-cash system generates better gains because of volatility of our system. Supposing a dollar's investment halves and then doubles. A buy and hold system made no profit. But a system which invests only 50% of available cash in investment makes some money. You first 100 dollars invested only 50% becomes (after first day when value halved) 50+25=75. Now on second day you invest 37.5 keeping aside 37.5. The invested amount doubled and became 75. Now you have 75+37.5=112.5 and you have made 12.5% whereas buy-and-hold would have made no gain. Of course, commission cost prevents you from making too frequent adjustment to size. Amd one need to take a total view of all factors. In this Covered Warrant system I am paper trading sizes are small. Commission of 25 GBP on 625 GBP investment is 4 percent. We have to make this money from price moves in shares we bought over and above the bid-offer spread loss. This is huge.