Correctly hedging yourself while taking overnights

Discussion in 'Trading' started by john_nyc_trader, Oct 27, 2003.

  1. As a trader who primarily closes positions at the end of the day, overnights are a non-issue to me. However, one recent trading day, the firm's (which will remain nameless) trading system went down, leaving me overnight with a sizeable (2500) short position in CTX.

    Luckliy, I came out unscathed, however, technlogical problems always arise, and my aim of this thread is to discuss practical ways in hedging yourself in situations such as this.

    For example, in hindsight, I probably should have hedged myself by going <b> long the SPY's. </b> However, the $64,000 question is, which method is better and more effective?

    A) Go long 2500 SPY's, share for share

    B) Go long by equal dollar amounts, i.e. CTX (2500 * $95.01) = $237,525; $237,525/$103.63=2292 SPY shares

    C) Not going long the SPY's at all; going long something more correlated. i.e. KBH, RYL, TOL etc...

    I would love to hear some your experiences & methods.

    John
     
  2. I assume you're talking about buying the hedge in an alternate account with a backup broker?

    Takes a lot of spare margin to buy 2500 shares of SPY. The e-mini S&P future contract might give you the market risk hedge you need for a lot less money.
     
  3. shaq48

    shaq48

    spys don't work so well as a hedge for the builders..and if you have to go in after hours and buy 2500 shares of a homebulider your going to move the after hours market all by your lonesome. The builders are all trading at 52 weeks highs or within a whisper of them..not even 52 week highs , All time highs. The spy will be indicated up 20 cents in the morning and a short gets nervous and puts in a buy on the open for a measley 5k and they open CTX or BZH up a buck. These stocks all have momemtum now and the only sellers are shorts who cover quick. I guess your making money shorting these or you wouldn't be doing it. No good hedge for these babys right now so be very careful shorting them.
     
  4. Shaq, if it wasn't for the system outage, trust me, I wouldn't have been short overnight. My question really is not focused on homebuilders, however, it really questions appropriate hedging methods for stocks that are not a major component of an index.

    For example, if I was short MSFt overnight, I would go long the QQQ's because it is a major component of the QQQ's, right? However, how would you hedge yourself if you were long HDI (Harley Davidson)? would you short the DIA or a stock that is remotely close to what HDI does (is there something out there?)

    confused :confused:

    John
     
  5. 'Back in the day' it was easy....take home something weak for sometihng strong....and ratio it according the adv/dec etc for the overall market. Anybody keep track now if that works? I doubt it.
     
  6. shaq48

    shaq48

    There are stocks that are just plain hard to hedge ...AZO for example is tough to hedge. So if holding overnight just be aware that some stocks just don't have a great hedge.
     
  7. Long another homebuilder to hedge. Homebuilders often move indepedent of the SPYs. Best way is of course to have a backup account with different brokerage and countwer the position with the same stock.

    Hope it opens lower.
     
  8. JayS

    JayS

    If your are able to hedge from another acct with the SPY or a similar stock thats correlated why not just offset the position in the other account that your able to do these "hedges" from.
     
  9. ler

    ler

    Or how about volatility matched hedging?

    i.e. the 30-day historical volatility of SPY is 13.68%, CTX is 29.50%

    You can go through the math to take annualized numbers to daily but its just the ratio that counts.
    I solve:

    $spy * HV30spy * SHRspy = $ctx * HV30ctx * SHR ctx

    and come up with 4943 SPY to match, call it an even 5000

    (the HV30's are close-to-close numbers, to put a really fine point on it you'd probably want close-to-open data for overnights)

    This approach matches risk, but doesn't consider your overnight cost of capital -- then you might just rather match the $$
     
  10. Has this approach been successful in your trading?
     
    #10     Oct 28, 2003