Swing Trading System - Calculating Buying Power - I Need HELP PLEASE!!!

Discussion in 'Risk Management' started by trendman, Apr 9, 2017.

  1. trendman

    trendman

    Thanks in advance. I have a swing trading system and I was asked by a hedge fund interested in running it how much buying power I need. I'm having a hard time calculating the # I should give them. Any advise would be appreciated.

    Normally I risk 1.5% of the portfolio on any trade, so if I have $1MM AUM, my risk on a trade would be $15,000. If my stop is 3 points, I'd buy 5000 shares. Fairly straight forward.

    Here is my issue. I have had times where I have 20 positions opened and times with as little as 5 positions opened. Its up to the algo and the market. Stock prices range form $5 to $900. That's quite a variance.

    Any thoughts on how I can calculate a buying power #? Futures are so much easier (and that's where I come from). X dollars = Y contracts. Any help is greatly appreciated as I need to have a number by tomorrow.
     
  2. JackRab

    JackRab

    So you would buy 5000 shares, no matter if it's trading at 5 or 900?
    That doesn't make sense. You can't just say 3 points/ticks.... it needs to be related to price...

    If that 3 points would be say 3%?... than you would use 500k on a single trade.
     
  3. trendman

    trendman

    The 3 pt stop was just an example. Each trade has a different stop point wise. I might risk 4% of the stock price on 1 trade and only 2.5% of the stock price on another trade, but no matter what, I'd only risk 1.5% of the portfolio on any 1 trade. Make more sense?


    $1MM Account with a limit of 1.5% of the portfolio value risked per trade
    Buy 100 - Stop 95 - 3000 shares
    Buy 50 - Stop 48 - 7500 shares

    Each trades price risk fluctuates but the risk to the portfolio % wise doesn't.
     
  4. algofy

    algofy

    20 positions you are going to be heavily correlated and much greater than 1.5% risk. Something smells a bit fishy with this whole post if I must be honest.
     
    HolyMoly, tommcginnis and Visaria like this.
  5. trendman

    trendman

    Your not reading it correctly. 1.5% risk per trade. On a $1MM portfolio, I will risk $15,000 per trade. The price change (points and % wise) changes from trade to trade. So if I am risking 10% of the current price of a stock, I would take marketably less shares than if I was risking only 2% on a trade. I'm not sure what is fishy. Its a pretty straight forward question.

    Additionally, there is no way for you to know anything about the correlation based on the # of positions. I could be holding half shorts and half longs across 20 different sectors for all you know.
     
  6. Have you done any kind of backtesting showing how much money you needed "worst case" at any moment of time? With worst case I mean the time that you invested the highest amount of money in your positions?
     
  7. trendman

    trendman

    All forward testing and the max # of positions opened during the test was 7 and they were all sub $100 stocks. I have had as many as 25 positions opened over the past few decades using this system as part of my broader trading, so it cant easily be broken down. And when I had 7 positions, they were cheaper issues. If they were all GOOG's, my necessary BP goes up 400%, hence my quandary.

    So the "worst case" as you stated was too small.
     
  8. JackRab

    JackRab

    So what does that HF want to know? How much buying power you need...?

    So are they basically asking how much you want? Or whether it's scalable up to what level? Because if they would ask me how much buying power I need, I would say 100 mln... depending on whether it's scalable. Are you using leverage in your strategy? How easy can you move 1 mln or 5 mln in a single trade? If you can only trade 100k on average per position/trade based on the liquidity... then it sounds like that's what they want to know.

    If it's massively scalable, and you can trade 20 mln easily (for instance when it's ES) then that's the number. If they are serious, they want a serious number. If you personally don't want to lose more than 15k on a trade, they might be fine with 50k or 100k depending on the stats...

    So it's about what they really want and how the strategy works in scale I would say.
     
  9. Maybe you need to turn it around: instead of having a sum of money which can handle all possible situations, including the "worst case" ones, you could make the system such that it adapts to the available money. Is it able to do that? And then ask for the maximum amount of money it used during testing.
     
  10. trendman

    trendman

    Thanks all. We came to an agreement today. The accepted my number, which is the max amount the system cam handle, with a right of first refusal on my next 2 systems, which are in final testing now.

    Thanks for all the input.
     
    #10     Apr 10, 2017