The BTC Horse Racing Thread
-
@Jonathan-Jones
Yeah, I guess the risk you take on with a shared bank may be worth it for the early stages to grow your bank faster, particularly if starting off small. Could always split the bank up once you get to a stage where having a big drawdown (say 50%+) seems too scary..
-
Interesting isnt it? No doubt a shared bank is a much bigger risk due to a wayward strategy (particularly if you are introducing an untested strategy) but i think it comes down to the bank size. If you are building a small bank then you get much more benefit for a smaller investment from a shared bank. If you have a good bank already and you are after an income or retirement type bank building then a split bank might be what you want.
-
@Jonathan-Jones
I just did some analysis of my trades this month. Running 5 strategies with split equal banks I am around 14% up (not been an amazing month to be honest, two strategies are down). If you export every bet and assume 1% stake of a rolling total bank the return is 65% with a max drawdown of 30.25%. Quite interesting..
-
I wrote a small app to help me reconcile my results every day. I download the individual results and import them. By default it runs the shared bank but strategies are selectable so i can just turn ones off to get what a single one does (that would match the backtesters output directly). So for this the shared one is obviously already done. I just ran the others individually then added the daily returns in excel and worked out the DD.
The massive difference was what made me look at it in detail because it really surprised me. But as a simple summary i guess you coud say you either:
Get approx 5 times the profit for a third of the investment but put up with a considerable DD on the shared bank.
or
Get approx a fifth of the profit for about a third of the DD with the split bank.So do you make the bigger investment and play it safe (and accept lower returns) or invest a smaller sum and take the ride? If i had a big bank then its a no brainer but for a very small bank to grow then its a bit less clear cut.
-
@Dan-MacKinnon said in The BTC Horse Racing Thread:
@Jonathan-Jones really interesting. Just out of curiosity, how did you manage to track all 3 together?
Wow, that is some food for thought. That’s the question I was going to ask!
-
@Jonathan-Jones really interesting. Just out of curiosity, how did you manage to track all 3 together?
-
Got some numbers.
Its based on the 3 strategies i run (2 lay, 1 back). If i started with £99 total bank (just to make it easy) at the start of jan.With all 3 strats sharing that total bank and a stake size of 1% i would today have a balance of £7161. The biggest DD though would of been 49.59% of that total bank which would of at the time it happend been a monetary drop of £316 (but that max DD lasted for only 5 days, 6th day cleared it).
If i took that £99 and split it evenly so each strat is managing its own bank of £33 with the same 1% stake size then today i would have a balance of £421.74. The biggest DD was 16.64% which represented a monetary drop of £35.
Just as an extra if i was just wanting to play with penuts and started a shared bank with £33 (so i really dont give a crap if i lose it, just a round of beers really) then by same rules i would have £2386. Obviously same %DD of 49.59% and a monetary drop of £105.
So not sure what that says really. Obviously splitting the bank is miles safer but at what point does being safe trump making some progress? If you have a fair bit of money then fair enough but if starting small then sharing the tiny bank you dont care about losing thats the same size as one of the splits is 5 times the profits for a third of the investment.
-
@Jonathan-Jones
If you pull together any numbers to compare the two approaches I'd definitely be keen to take a look!
-
Was thinking that way so its seems safer but the difference in account growth between the methods is just massive. might run some numbers to get a proper comparison.
-
Hey guys, I am trying to use BFBM to automate a strategy that is based on BSP. It's fine for a strategy where I don't have any limits on the price, but I have a lay strategy where I want to implement a maximum odds.
The betfair interface allows you to do this by clicking 'Set SP odds limit' which then gives you a field to input the minimum odds for back bets or maximum odds for lay bets, but I cannot get this to work in BFFM!
I have gone through this thread in their forum but it is still not working:
https://www.bfbotmanager.com/phpbb3/viewtopic.php?t=3243
Any ideas?
-
@Jonathan-Jones said in The BTC Horse Racing Thread:
@Martin
Think so. I have tried InPlay stop losses on horses and they just dont do what you want. You either set it as per the strategy and they just dont match or you set them for a miles worse price and every loss is bigger than it should be.The SL on this strat just looks like it smooths the PL out a bit. Think the difference in total PL is only about 40 quid.
When you remove the stop loss it doesn’t seem to have performed that well in the last 6 months, which would be a concern.
-
@Jonathan-Jones I split the overall bank between the various strategies I am using to reduce risk of significant drawdown. Presumably if you set it to the shared bank/overall bank then you could have a pretty big reduction to the overall bank on a bad run, particularly if multiple strategies hit a rough patch?
-
Question for those of you who use BFBM in regards to the split banks (or anyone who uses any similar functionality really).
Do you use it with multiple strategies so each strategy just manages a fraction of the overal bank or do you just use the whole bank shared so all the strategies share the profits/losses?
I am guessing the first is the safer option but it looks like there is an absolutely massive difference in the rate of account growth (shared bank being a lot more profitable).
-
@Dan-MacKinnon yeah assume it would work fine on prerace as moves are smaller
-
@Martin I tried a similar strategy a while back when the horse racing software was first released. It used a pre-off SL and even when I moved the trigger further apart, as Jonathan said, it always seemed to match on "winning" trades and didn't when I wanted it to! I gave up with SL and think it's more of an insurance unless you're manually trading
-
@Martin
Think so. I have tried InPlay stop losses on horses and they just dont do what you want. You either set it as per the strategy and they just dont match or you set them for a miles worse price and every loss is bigger than it should be.The SL on this strat just looks like it smooths the PL out a bit. Think the difference in total PL is only about 40 quid.
-
All the time. I tried this one live when it first hit the forums. in terms of looking at the backtest and live results you cant compare them. No idea how this would perform live long term. Only thing for certain is it wont be anything like the backtest. I tried an inplay SL again recently in one of my strategies. Only ran it for a couple of days because its simply not worth it. People need to remember a stop loss isnt a standing order. Its placed when the conditions for it are met so its subject to the inplay delay and because of the way prices move there are some universal truths about stop losses activated inplay.
- A stop losss you might not want matched because price then reverses again and goes in your favour will ALWAYS be matched.
- A stop loss you really need matched because price is storming against you will NEVER be matched.
You can with a stop loss trigger at a value but set the price miles out (worse price) then its probably going to get matched but you cant model that in the backtest software. Changing the stoploss value in the backtest doesnt model this.
This particular strategy looks pretty good without the stoploss though. Its just a bit more of a jiggly line.
-
@Martin said in The BTC Horse Racing Thread:
@John-Folan interesting what you were saying about stop losses sometimes being blown through, how far would that happen?
Depends on the race liquidity plus how quick the horse fades etc. It does happen a lot.
-
