Bollinger on Hurst’s Envelopes
The next major development came in 1970 when J. M. Hurst published The Profit Magic of Stock Transaction Timing.
Hurst’s interest was in cycles,and he used ‘‘constant width curvilinear channels’’ to clarify the cyclic patterns in stocks. His approach was to use multiple hand-drawn envelopes (see Figure 6.6) that related
to the various cyclic components of price action.
The envelopes nested inside one another,often becoming congruent at major turning points. In the back of his book,he gave some broad hints at how this process might be mechanized (see Figure 6.7) but the examples presented in the text appear to be hand-drawn.
We suspect that the concepts were beyond the technology then commonly available. In the years since,numerous attempts have been made to systematize Hurst’s work,but we are not aware of
any successful results.
The development path of trading bands gets a bit murky here, and credit is hard to assign. In the next phase,inter est seems to have broadened,and several analysts appear to have been working on similar ideas at the same time. The main technique employed in this phase was to shift a moving average in a parallel manner up and down to form bands around price (Figure 6.8).
The offset was typically a number of points or a percentage of the average.
See,for example,Table 6.2. Hurst had clearly favored the use of moving averages in his book,but we think the idea of shifting the averages by some mechanical means came later, perhaps in the early 1970s. The problems of this approach became apparent immediately.
First,the width had to be determined empirically on an issue-by-issue basis.
Second,even having done that, the widths needed adjustment over time. Thus, while percentage or point bands did provide useful definitions of high and low for traders, they were hard to use and involved considerable guesswork on the part of the user.