Brain-training tasks related to trading in one way or another. Theorist, game theory, etc. - page 13

[Alexey Subbotin]

timbo:

I ran 10,000 simulations for 28% in MATLAB, here is a histogram of the lifetime of this strategy, i.e. before it is lost. The vast majority of cases (90%) are lost before the 100th trade. Very few people last longer. I.e. failure is guaranteed.

purely out of curiosity - does this include the additional cost of commissions/spreads?

[TVA_11]

Reshetov:

b - potential gain in money / potential loss in money = 3 / 2 = 1.5

((1.5 + 1) *0.5 - 1) / 1.5 = 0.16666666666666666666666666666667

((2+1)*0.5-1)/2

=

0,25

Then the result should be 0.25

but in practice, it turns out to be 0.28.

Usual eagle's game - we win one and lose one.

Here we win two, lose one.

b - 2/1=2

[TVA_11]

alsu:
Just for the sake of argument - does this include additional expenses on commissions / spreads?

100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100
100	0,5	50	200
200	0,5	100	100

With 50% of deposit and winnings against one coin bet 2 and loss of 1 coin bet with 50% probability - there will be a flat.

If betting more than 50% - then it is a flat. Less than 50% - capital gain. Maximum at 28 per cent. Not 25%.

[[Deleted]]

alsu:
purely as a matter of interest - does this include extra charges for commissions/spreads?

No extra costs, everything is strictly on condition.

[[Deleted]]

TVA_11:

With 50% of deposit and winnings against one coin bet 2 and losing 1 coin bet with 50% probability - there will be a flat.

If betting more than 50% then it is a flat. Less than 50% - capital gain. Maximum at 28 per cent. Not 25%.

Yeah, well... And Kelly's an underachiever... You've turned the whole maths thing upside down...

[[Deleted]]

TVA_11:

With 50% of deposit and winnings against one coin bet 2 and losing 1 coin bet with 50% probability - there will be a flat.

If betting more than 50% then it is a flat. Less than 50% - capital gain. Maximum at 28 per cent. Not 25%.

For anyone interested, just in case, let me clarify that this is the nonsense of a dumb triplet. The correct figures and formulas are given above.

[TVA_11]

100	0,25	25	150
150	0,25	37,5	112,5
112,5	0,25	28,125	168,75
168,75	0,25	42,1875	126,5625
126,5625	0,25	31,64063	189,8438
189,8438	0,25	47,46094	142,3828
142,3828	0,25	35,5957	213,5742
213,5742	0,25	53,39355	160,1807
160,1807	0,25	40,04517	240,271
240,271	0,25	60,06775	180,2032

Sorry, recalculated Excel.

[TVA_11]

Capital appreciation rate per transaction =

12,5/2 = 6,25 %.

[Yury Reshetov]

TVA_11:

Capital appreciation rate per transaction =

12,5/2 = 6,25 %.

This is a geometric progression, not an arithmetic progression. Therefore, the return should be counted as a geometric progression:

Two coin flips: 1.5 * 0.75 = 1.125, i.e. by (1.125 - 1) * 100% = 12.5%

One flip of a coin: (1.5 * 0.75)^ 0.5 = 1.06066, i.e. a geometric mean gain of 6.066%

[Rossi]

Problem: the indicator makes a correct entry 75% of the time. ( i.e. takes profit from 0 to 3% ) What should be the optimal lot size, e.g. to avoid 20% drawdown in 100 trades? And what should I do with lot size in case of failure ( 2% stop loss exit) - should I multiply it by coefficient or leave it unchanged?

there are slow draining EAs, e.g. on MA crossing... will varying the lot size change the draining rate?