A Sports Hedge Fund?

I was alerted to the creation of a new sports hedge fund by Pyckio. Not for the first time has a company attempted to put their sports trading activities on a similar footing to a hedge fund, Centaur Galileo being the most prominent, in that it soon collapsed.

The most likely reason for the collapse of Centaur was market capacity, with too much fund money being traded in a market such that the fund became the market and eroded the fund's edge.

As I have repeatedly said on this website and in Betfair Trading Techniques, too many novice traders have some initial success, usually through variance (luck) and then run a compounding spreadsheet to forecast their future wealth.

The upshot of the erroneous compounding exercise is that the novice imagines they will soon be a millionaire. In reality, the size of the novice's trades will have to increase such that there will not be enough money on the other side of the spread to match the novice's trades.

A trader must always take into consideration the amount of money they can match against the edge they have found. Put too much money into the market and you can easily end up trading against yourself if the spread moves against your edge.

Pyckio is a tipster aggregation website and such websites can inadvertently (or otherwise) become the victims of survivorship bias. Joseph Buchdahl in Squares & Sharps, Suckers & Sharks has performed extensive research on tipsters and tipster aggregation websites, revealing that most winning tipsters are merely lucky and some aggregators are fraudulent.

Assuming Pyckio is reputable, and there is no reason to doubt that if Buchdahl advertises Pyckio's fund without query (Buchdahl is actually a partner at Pyckio), then any attempt at creating a fund using winning tipsters will run into two problems.

The first problem relates to edge. If tipsters are lucky then they will mean revert and anyone investing in the fund will see their capital gains eroded by the loss. Any fund administration charges will further erode the fund such that investors will lose money in the long run.

The second problem relates to market capacity in that every winner needs a loser to compensate them. Pyckio will be limited to how much money it can put into a market by the amount of losing trades that are entered into the market.

A sports betting market is in no way comparable to a financial market. A single betting market might see a few million pounds changing hands. Compare that to the trillions in financial trading. There is much more capacity for a financial hedge fund to manoeuvre within.

If Pyckio is successful in its enterprise then it will become the victim of that success. Why bother betting or trading when you can get a handsome return from Pyckio's fund? The more people that join Pyckio's fund the less edge it will have over the market because there will be fewer losers.

Losing traders will join Pyckio's fund and will no longer be losers. You cannot have a market where 100% of participants are winners and so Pyckio's edge will dwindle. This means that Pyckio will have to limit the number of people that join the fund and fund members will have to pay a premium for that exclusivity. Again, eroding any edge a fund member may think they have.

Pyckio are walking on a tight rope with this venture. I wish Pyckio well with their fund but more so anyone investing in such a fund.

8 comments:

  1. "A trader must always take into consideration the amount of money they can match against the edge they have found. Put too much money into the market and you can easily end up trading against yourself if the spread moves against your edge."

    Hi James.

    This is a sentence that stuck on my mind. Do you have any method to set your stake's size according to the market volume? Or do you use any specific algorithm based on some kind of metric (volume traded in the last X minutes, available to back/lay sizes, etc)?

    Thanks!
    Pedro P.

    ReplyDelete
    Replies
    1. A good question. Betfair Trading Techniques discusses the problem of trade sizing in detail. I will write an article about some aspects in a future article.

      Delete
    2. Yes, I own that excelent book of yours! I'm talking specifically about the Market Capacity section (p71).

      Sometimes I find it a bit hard how appropriately calculate the max size of my bets according to what the market is offering in terms of volume. Sometimes I feel that I should have bet more and sometimes that I should have bet less, especially when the market goes against my expectation eheh. All in all I guess I'm not happy with the way I calculate the bet size for these kind of markets.

      Thanks!

      Delete
    3. I will keep those comments in mind when I write the article.

      Delete
  2. Hi, I’m Daniel Mateos, CEO of Pyckio. Thank you for your article. I would like to make some comments and answer your concerns.

    First of all, about the Centaur Galileo fund, they told investors that they lost it all due to "sheer bad luck". I think this is a sign that the managment wasn’t too professional precisely.

    We have real professional experience in financial markets and we’ve been profesional bettors. We’ve also Ph.Ds in Economics and experts in statiscical models. We know well how the betting market works and we are conscious of the liquidity issues. Of course, our expertise per se doesn’t guarantees that the fund will be succesfull. I just want to make clear that we’ve decided to embark on this project knowing the reality and the constraints.

    About the tipsters survivorship bias, we are also fully aware of it. We know how difficult is to find genuine long term winners but we’re also sure that working with Pinnacle + some exchanges + asian bookies, we can achieve a low aggregate margin and therefore be capable of achieving a positive low expected aggregate yield in thousands of bets every year. About our risk system, the weight of every individual bet will be negligible.

    To cope with the survivorship bias, we’ll set up a 2nd qualitative filter, that will be mainly personal interviews with the potential analysts that will feed the fund. From our experience, when you speak with a tipster, you can't never assure if he will succeed in the long term but with some of them you can certainly assure that they won’t. Tipsters who do not understand that this business isn’t about finding winners but about detecting wrong odds are doomed to failure. We are pretty sure there are tispters out there who are genunine long term winners. They aren’t the norm obviously but they exist.
    Our intentions are to have some of the best analysts (from Pyckio and also from outside) in as many as competitions/leagues as possible. The Fund will be highly diversified (analysis, sports and competitions) to cope with the liquidity issue as much as possible. As you write, as long as the fund gets bigger the liquidity constraints will be higher. We know this reality and we’re not ambitioning a 100 Mill. € fund. We’ll be monitoring closely the market capacity to stop accepting subscriptions when we notice the market cannot longer accept more money coming in. Moreover, as long as we’ve got more funds to manage we’ll be adding additional sources of alpha such as live betting strategies. We might also incorporate more quantitative analysts to our team. Any strategy (live or prematch) that is presented to us and passes our due digillence analysis might also be incorporated. In exchange and leveraging on the potential profits we can generate, we’ll offer the “alpha provider” a percentage of the potential gains. Therefore, anyone could be able to leverage on our fund as a way to exploit a winner betting system.

    We know there are big syndicates out there who win money betting time and again. There is one in IOM with around 30 people working, what suggests there is quite posible to achieve a positive return if things are done very well. At the end, our task will be constantly find inefficient odds everywhere, be it through the analysts/tipsters we’ll collaborate with, via our in-house generated systems or with external ones. About in-house generated systems we’ve got a great set of Big Data (> 5 Mill. Picks submitted at real Pinnacle odds) that we can take advantage of. We know bettors biases and we know where they are most likely to win and where they are most likely to lose. That is, we know where inefficiencies are more likely to be found.

    To sum up, we’re quite confident we’ll be able to deal with the market capacity issue and hope to be finding new sources of “real positive expected yield” as we get bigger. Shouldn’t we find them, we’ll stop accepting new subscriptions so that the fund performance won’t be accepted.

    Kind regards

    ReplyDelete
  3. Hi James, I wrote my comment too quickly and made some erratas. I've just fixed them. Could you please change it? You can find it below. Thank you.



    Hi, I’m Daniel Mateos, CEO of Pyckio. Thank you for your article. I would like to make some comments and answer your concerns.

    First of all, about the Centaur Galileo fund, they told investors that they lost it all due to "sheer bad luck". I think this is a sign that the managment wasn’t too professional precisely.

    We have real professional experience in financial markets and we’ve been profesional bettors. We’ve also Ph.Ds in Economics and experts in statiscical models. We know well how the betting market works and we are aware of the liquidity issues. Of course, our expertise per se doesn’t guarantees that the fund will be succesfull. I just want to make clear that we’ve decided to embark on this project knowing the reality and its constraints.

    About the tipsters survivorship bias, we are also fully aware of it. We know how difficult is to find genuine long term winners but we’re also sure that working with Pinnacle + some exchanges + asian bookies, we can be capable of achieving a positive low expected aggregate yield in thousands of bets every year. About our risk system, the weight of every individual bets will be negligible.

    To cope with the survivorship bias, we’ll set up a 2nd qualitative filter, that will be mainly personal interviews with the potential analysts that will feed the fund. From our experience, when you speak with a tipster, you can't never assure if he will succeed in the long term but with some of them you can certainly assure that they won’t. Tipsters who do not understand that this business isn’t about finding winners but about detecting wrong odds are doomed to failure. We are pretty sure there are tispters out there who are genunine long term winners. They aren’t the norm obviously but they exist.
    Our intentions are to have some of the best analysts (from Pyckio and also from outside) in as many competitions/leagues as possible. The Fund will be highly diversified (analysis, sports and competitions) to cope with the liquidity issue as much as possible. As you write, as long as the fund gets bigger the liquidity constraints will be higher. We know this reality and we’re not ambitioning a 100 Mill. € fund. We’ll be monitoring closely the market capacity to stop accepting subscriptions when we notice the market cannot longer accept more money coming in. Moreover, as long as we’ve got more funds to manage we’ll be adding additional sources of alpha such as live betting strategies. We might also incorporate more quantitative analysts to our team. Any strategy (live or prematch) that is presented to us and passes our due digillence analysis might also be incorporated. In exchange and leveraging on the potential profits we can generate, we’ll offer the “alpha provider” a percentage of the potential gains. Therefore, anyone could be able to leverage on our fund as a way to exploit a winner betting system.

    We know there are big syndicates out there who win money betting time and again. There is one in IOM with around 30 people working, what suggests it's quite posible to achieve a positive return if things are done very well. At the end, our task will be to find inefficient odds everywhere, be it through analysts/tipsters, via our in-house generated systems or with external ones. About in-house generated systems we’ve got a great set of Big Data (> 5 Mill. Picks submitted at real Pinnacle odds) that we can take advantage of. We know bettors biases and we know where they are most likely to win and where they are most likely to lose. That is, we know where inefficiencies are more likely to be found.

    To sum up, we’re quite confident we’ll be able to deal with the market capacity issue and we hope to find new sources of “real positive expected yield” as we get bigger. Shouldn’t we find them, we’d stop accepting new subscriptions so that the fund performance won’t be affected.

    Kind regards

    ReplyDelete
    Replies
    1. Sorry for publishing this comment late but I never received an email from the server telling me that it had arrived.

      Delete
  4. Well if I was ever interested in such a scheme (which I am not) then this reply by the "CEO" has certainly confirmed my opinion about this one. What a lot of tosh he writes (and poorly to boot). Yes, let me give you my profitable betting strategy so I can become an "alpha provider" so that we can leverage that edge out of existence. Thanks for the thoughtful and thought provoking site James.

    ReplyDelete