A Different Form of Trading

[kermit345]

Hi All,

I've been reading the forum quite a bit lately but haven't posted much besides the odd comment here and there. Markets have just been too frustrating for me in more recent times and i've been concentrating on a few others things (Work/Life) so therefore haven't been able to contribute much to a lot of the discussion.

However something i've also been working on is approaches to gambling (i.e. sports and racing betting). Majority of you probably find this hilarious but I figured i'd post here as some people may be intrigued and I might get some quality input from people. First i'll give some background as to what i've done and then some detail on what im looking to find/achieve in this forum.

Over the past 2-3 months i've developed on excel spreadsheet that allows me to complete my own ratings on Australian horse races, essentially providing in my view who I think the favourites are etc etc. Basically i've refined this over time and found a way to bet on Australian horse racing profitably. However unlike most gamblers i've also approached this with a share traders mindset as well looking for the best risk/reward combination and always searching for value (betting on horses at high odds when my system rates them as a top chance - going against the crowd somewhat).

I've used share trading statistics that i've found in a Tech/a thread from ages ago to help gauge how my betting approach is going. The following is an output I posted on a horse betting forum some weeks ago (it's actually more up-to-date now but as i'm not home I cant post currently but will tonight).



As you can see its doing quite well, but I was curious as to what other statistics, charts etc people would look at to understand how healthy the system actually is. Over the last few nights i've added charts showing the maximum drawdown on my bankroll and also how many bets have been x standard deviations away from my mean to see if outliers are carrying the bulk of my profits (and hence the system isn't actually that good).

All questions, comments are welcome. Hopefully this thread doesn't get closed down as i'm trying to take a traders approach to the betting of the system and I legitimately think a share traders mindset can be applied, and very well i'll add.

So any comments and suggestions about statistics or charts that I may be missing to help me gauge and monitor the health of my trading (betting) wold be greatly apprecaited.

Cheers

 

[prawn_86]

Some very impressive stats there Kermit, feel free to start an ASF betting syndicate

Seriously though, i think you probably have it covered stats wise, and depnding what you intend to do with this you could take out some funds/profits out now and let the system run just building profit

Are these trades automated or do you place them yourself? Is it just betting on win/places or various other bet options?

 

[barney]

Hi All,

Majority of you probably find this hilarious

So any comments and suggestions about statistics or charts that I may be missing to help me gauge and monitor the health of my trading (betting) wold be greatly apprecaited.

Cheers

Hi Kermit, I don't punt on the ponies anymore, but those statistics are definitely no laughing matter

If you are able to produce results even half that good over an extended period, your system could be worth its weight in gold.

If you are able to keep the results anything like the current ones for any length of time ...... one word springs to mind ...... "compounding" ....... you could end up owning the local race club

 

[kermit345]

Hi guys,

Thanks for the kind words, me and a friend are placing the bets manually at the moment based on whether the spreadsheet tells us to bet or not. I expect the results to become a bit more realistic over the long-term as I really don't think the current results can be sustained, but if they are then i'll be more than happy to take the ride.

prawn - this is based on win bets only on my top rated horse which through monitoring various other stats i've found can even be bet profitably without any additional filters. The system I use has some filters in place to improve the results (i.e. Slow/Heavy tracks show poor results, Thursdays seem to do poorly and country QLD races seem to perform poorly). I've found that my top 5 rated horses win approx 77% of the time which is pretty much spot on with the top 5 favourites from any race are expected to win this amount. The advantage my system has is that my top 5 may not be favourites and hence im hitting the same amount of winners as favourites but achieving better results through greater odds.

Anyhow thats the betting side of it. I'll be home on my lunch break in an hour, i'll post up-to-date results and the max-drawdown / Standard Deviation charts for those interested.

By the way the max drawdown so far is 18% based on betting 2% units of my bankroll, I assume this is quite good in comparison to share trading max drawdowns?

 

[CanOz]

Very interesting Kermit...I've had discussions with my old mates back in Australia before about the horses and how we might apply some money management to a system. Well you've done it so far....

CanOz

 

[kid hustlr]

Looks very interesting

 

[kermit345]

As promised. Updated stats and charts:



The following charts are in the following order:
1) Bankroll
2) Max Drawdown
3) Standard Deviation Measurements

 

[CanOz]

Even got a scatter gram, well done. Nice work on the stats Kermit!

 

[Trembling Hand]

Thats a beautiful scatter chart.

 

[sinner]

Good stuff kermit, there is a lot of scientific work which backs up your findings!

http://en.wikipedia.org/wiki/Favourite-longshot_bias

 

[kermit345]

haha I added the scatter chart last night as I was concerned that my outlier bets were not showing the true picture of how it performs, but I think its shown that while there are some outliers the majority of my winners are around the mean or within 1 standard deviation which means my stats shouldn't be overly optimistic.

Thanks for the interest guys, I didn't think many here would be interested haha. As i've said if you have any suggestions on other charts etc I could include please let me know. I'm thinking of building a spreadsheet for others to use for their betting as a lot of the guys were impressed and hadn't seen the kind of stats I was showing before.

Sinner - what made it click about a month or so ago for me was reading the following article. It's regarding a guy called Allan Woods who made millions betting on horses in Japan based on finding value (i.e. horse is at $5 odds but his research shows its a $2.50 prospect). He then adjusted his staking size depending on how big this value was (overlay's and underlay's). He had a full team who analysed a lot more aspects then what my little spreadsheet can do, but its basically the same idea. You may find it an interesting read:

Alan Woods - http://www.tonywilson.com.au/writing/alanwoods.php

Some of you may be surprised how much horse betting can correlate with technical analysis. There are programs that can be fully automated (much like share trading) to bet on horses based on the same indicators as share trading. Also the same as shares, you can back or lay (long or short) a horse to win on a price going up or down. I could find more detail if anyone is interested.

 

[sinner]

Sinner - what made it click about a month or so ago for me was reading the following article. It's regarding a guy called Allan Woods who made millions betting on horses in Japan based on finding value (i.e. horse is at $5 odds but his research shows its a $2.50 prospect). He then adjusted his staking size depending on how big this value was (overlay's and underlay's). He had a full team who analysed a lot more aspects then what my little spreadsheet can do, but its basically the same idea. You may find it an interesting read:

Alan Woods - http://www.tonywilson.com.au/writing/alanwoods.php

Totally dig it. Here's one for you as thanks for your link

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1420356

Some of you may be surprised how much horse betting can correlate with technical analysis.

Not me!

 

[jancha]

Thats a very good win strike rate 44%. Do you do your own rating and are your bets fixed price and if so what is your cut off betting eg 5/1?
I often look at the 100 raters at odds with good results sometimes especially with multiple betting.

 

[prawn_86]

Slightly off topic, but another good 'system' is that of tennis multi's. I have not actually tested it with stats etc, but loading up on multi's (say 5 - 8) matches and always picking the favourite in each match, seems to bring in more $ won than $ lost. It does suffer from large drawdown periods though as the majority of bets are losers

 

[kermit345]

jancha - the ratings are done myself and are fully automated through excel using web queries. So basically I put in the date, venue, race no. and hit submit and my spreadsheet spits out my ratings and if/what to bet. The tracking of my bets is using Unitab odds, but if you were to place with Betfair or a website that offers the 'best of all totes' type service then the results would be even better then what i've provided. I have no limit on the odds I take as if my system says to bet on a horse (i.e. I rate it highly) but the odds are $10-1 that is a perfect scenario as i'm getting massive value on something i rate highly compared to the public. It's a bit like buying a stock at $4.00 that may potentially reach $5.00, but the market (public) think its already fully priced at $4.00 - i'm achieving that value gap.

prawn_86 - I haven't played much with tennis multi's although I have heard it can be done successfully. I've gradually became less and less interested in multi betting and exotics - the returns seems to be great but the risks become out of control quite quickly. In the previous NBA and NHL seasons in America I experimented with using various statistics for those sports to help me decide on if/when to make particular bets. It worked to some extent but not quite as well as i'd hoped and was quite labour intensive (not automated). I'm hoping that for the seasons coming up I can get some automated spreadsheets in place and give it another shot. I believe that following statistics can give you an edge against the general public if your selective in where you decide to exert that edge - it isn't available in all games and sports betting requires a LOT of discipline, often its too easy to waste $10 here and $5 there and before you know it your bankroll has dwindled to nothing because you haven't followed your own systems.

sinner - Thanks a lot for that link, had a quick browse at work but looks quite long so didn't get a chance to read in depth. I was lucky enough to have the opportunity through work to listen to Ken French speak regarding market returns and using value/small caps to achieve superior returns to the index. He went into a lot more depth regarding all of it but also explained it in a way that really opened my eyes. The thing that still sticks with me today is that it takes approximately 65 years of outperformance by an individual (i.e. fund manager) before they are considered to be able to beat the market regularly. There was an equation that Fama/French used to back this up not that I can remember it now, but basically this means there are a handful of people worldwide who can claim to beat the market as stock pickers regularly. The research Fama/French have conducted and what was put infront of me at his presentation was amazing, made me feel comfortable with my approach of value investing.

Note he was more talking about fund managers vs the market, not really individuals trying to beat the market.

 

[JLM Financial]

A large number of my clients have used expert advisers in the past. Like any automated system they will only work for a certain period of time. Once any market inefficiencies are realised they are quickly corrected.
Our firm uses a combination of fundamental and technical analysis for our recommendations.

 

[CanOz]

A large number of my clients have used expert advisers in the past. Like any automated system they will only work for a certain period of time. Once any market inefficiencies are realised they are quickly corrected.
Our firm uses a combination of fundamental and technical analysis for our recommendations.

Mate, they're talking about horses....:horse:

If you would like to advertise here, i suggest you contact the administrator, otherwise post something of value.

CanOz

 

[craft]

Hi Kermit

Nice stats.

Systematic betting certainly seems to have worked for Zeljko Ranogajec and David Walsh.
Have you been following the court case re tax on betting when conducted in a business like manner?
How are you going to approach the tax issue?

 

[Trembling Hand]

Mate, they're talking about horses....:horse:

Nice use of the smilies.

 

[peter2]

This post is a bit late due to the fact that those results just blew me away. It's taken a while to put the pieces back together.
Did I read it correctly, W% 41% with average odds $8. Results like this are fantastic and much too good to last.
The W% is too high for one bet/race and the average odds are too high for multiple bets/race.

Press your luck while the edge lasts, as I know it won't.