Cost normalization

Discussion in 'Trading' started by kmiklas, Sep 21, 2021.

  1. kmiklas

    kmiklas

    Brethren:

    I want to normalize risk across all my trades. In other words, take on the same amount of risk for each trade, despite different share costs and volumes.

    I've been adjusting my share count to achieve this, but have doubts. For example, I deploy about $10,000 on each position (minor differences to avoid fractional shares):

    Code:
    Apple: 71 AAPL @ 141.70 = $10,060.7
    U.S. Steel: 469 X @ 21.32 = $9,999.08
    Bank of America: 255 BAC @ 39.28 = $10,016.4
    I feel like I'm missing something... like I should be doing a VWAP, or because the share prices are different, it changes my risk levels. A 1% change in AAPL share price is very different that a 1% move in BAC price.

    - Nagging at me is the percentage move measure: if AAPL jumps $3, that's a 2.12% move; but if BAC jumps $3, that's a 7.64% move.
    - By setting the same total cost, does this also normalize risk?
    - If not, how can I properly normalize risk across different instruments?

    Thanks
    Keith
     
  2. deaddog

    deaddog

    I risk a percentage of my account.
    I take the difference between my entry and my stop and divide that into whatever % I'm willing to risk and that becomes my position size.
    My risk is constant not the position size.
     
    persistence likes this.
  3. Simple, I like that. A universal %.
    The simplest methods and explanations...tend to work out in the end.

    Keith, you like to over-complicate your trading and understanding and methodology and approach.
    Sometimes, it pays to be dumb -- or common and basic and general.
     
  4. 2rosy

    2rosy

    kmiklas likes this.
  5. ph1l

    ph1l

    According to

    do
    upload_2021-9-21_16-9-36.png
    In other words, take a proportion of the account value and divide by a volatility measure that has a dollars per share value. The result of dollars divided by (dollars per share) == number of shares to normalize risk.
     
    kmiklas likes this.
  6. userque

    userque

    You could normalize to the average percentage move of each instrument.

    With two instruments, for example; if the first instrument has an average daily move of 1%, and the second has an average of 2%, then you'd allocate half as much, percentagewise, into the second instrument. This figures the distribution; use another method to determine overall total allocation from available funds.
     
    tomorton and kmiklas like this.
  7. Snuskpelle

    Snuskpelle

    "In general" you want the position cash value (num of shares times share price) to be proportional to the inverse to volatility, whether that is measured by ATR, STD, or w/e. The devil is in the details I suppose, you could also include forward looking estimates of vol (from options).
     
    kmiklas likes this.
  8. MrMuppet

    MrMuppet

    Little by little you are getting annoying. He doesn't seem to be trading but allocating money in his portfolio.
     
    kmiklas likes this.
  9. MrMuppet

    MrMuppet

    Weight your portfolio by the individual stocks beta to the index. In other words, if a stock has a beta of 2 take half of the volume than of the stock that has beta of 1.

    A second option is by volatility weighting. Take the stocks annualized volatility and rebalance your holdings according to that. 30% vol gets half the allocation of a 15% volatility stock and so on.

    If you are in fact trading and you use stops then allocate according to a fixed percentage of your portfolio. 50$ stop gets half the allocation of a 25$ stop.
    However, this method is pretty complex since you have to determine the correct percentage. If your expected value is good, you can lose a bigger percentage of your portfolio per trade than if it is 50/50.

    Read the "Mathematics of Money Management" by Ralph Vince
     
    kmiklas likes this.