Using the nonadjustment method, you simply stop charting one contract
when it expires, or when you otherwise deem it justifiable to do so, and continue
to chart the next contract in line. This contract is the new front contract. Usually,
this coincides with when the market as a whole moves from one contract to the
next, resulting in an increase in the open interest for the new front contract sur-
passing the open interest for the old contract.
The main advantage with the nonadjusted time series is that it shows you exact-
ly how the front contract was traded at that particular time, with all trading levels and
price relationships intact and exactly as they once appeared in real life. Figure 6.6
illustrates what this looked like for the December contract on the S&P 500 index
during the crash of 1987. From a high of 333 on October 2, the market fell a total of
152 points, or $38,000 (152 * 250), to a low of 181 on October 20. In percentage
terms, this equals a drop of 45.6 percent (152 / 333) of the total market value.
The main disadvantage of the nonadjustment method is the differences in
price that frequently appear on the day for the roll; these distort your back-testing72 PART 1 How to Evaluate a System
FIGURE 6.6
S&P 500 December contracts—crash of 1987.