|
PureBytes Links
Trading Reference Links
|
At 12:26 AM 1/7/2004, Alex Matulich wrote:
>I was under the impression that ratio adjustment was simply a
>weighted average of the current contract and the next out-month
>contract, weighted according the position of the current date
>relative the the prior and next rollover dates. As such, the ratio
>adjusted data should be very close to the actual futures price no
>matter how far back in history you go.
>
>This is what I thought -- I have been using ratio-adjusted data
>as an approximation for the cash market as a means to generate
>signals far back in history. I use that for signals, and use the
>reverse-adjusted data for trading.
Not correct as I understand it.
I understand that ratio adjusting simply multiplies the old contract prices by a ratio to make them match the current contract prices.
For example, if the old contract price is 1050 and the new contract price is 1100, then the old prices are multiplied by 1100/1050.
This has the advantage that the adjusted price will never go negative, as it could do with fixed price adjustments.
It has the disadvantage that it distorts the bar-to-bar changes in prices. Thus, one tick on the S&P emini will shift from 0.2500 to 0.2619 for the just-expired contract. It gets worse as the contracts get older, obviously.
If your system uses bar-to-bar price changes (as most systems do), the fixed adjustment is better. You can add a fixed amount to all prices, if necessary, to prevent the price from going negative, (since these systems do not look at the absolute price - only the bar-to-bar changes).
Neither method will retain the actual contract price. The only way to retain the actual contract price is to use the unadjusted contracts.
Alternately, you could use an index as a proxy for the actual price level (but not for the bar-to-bar changes since the futures are noisier than the indices). This would be useful if, for example, you wanted to scale some stop to the actual level of the price.
Some time ago I wrote a paper on this topic that is posted on several web sites including:
<http://www.tradingrecipes.com/files/cntcontr.pdf>
<http://www.traders2traders.com/papers/backadjusting_futures_contracts.htm>
<http://www.ellokn.com/news/?CurrMonth=08&ID=6&Mode=Daily>
Bob Fulks
|