Futures traders can obtain important information about market trends by observing and analyzing the action of the open interest. The open interest is the number of contracts or options existing at the end of a trading session. You should note that every contract has a buyer and a seller who together make up one contract. The open interest should be monitored along with the price and volume action to determine whether an uptrend or downtrend is strong or weak. For more on basic open interest analysis, please see this previous lesson.
The Herrick Payoff Index (HPI) was developed by John Herrick and is the only well-known indicator that uses price, volume, and open interest. I was fortunate many years ago to witness a very lively debate between John Herrick and James Sibbett, the inventor of the Demand Index, about whose indicator was best. I personally never settled the debate for myself, as after using both for several years, I concluded that both were equally valuable. At the end of this lesson, I have included two weekly charts on commodities that are of current interest.
I find the HPI to be an excellent short- and long-term tool, and I recommend examining it on both a weekly and daily basis. HPI is simply a mathematical method of measuring the money flowing in or out of a commodity by computing the difference in dollar volume each day. The formula is quite complex (see below), and to my knowledge, is currently only available in Metastock software. The calculation requires two different inputs, the value of a one-cent move and a smoothing factor. For a one-cent move, I recommend 500 for financials and metals, and 250 for the rest of the commodities. The standard for the smoothing factor is ten, which relates to a ten-period moving average. The calculation of the Herrick Payoff Index (HPI) is:
Where:
The Herrick Payoff Index can be analyzed on many levels, and the simplest way is to determine whether it is above or below the zero line. If it is above zero, it shows that money is flowing into the market, therefore positive, while if it is below zero, it is negative. I have also found that by using a 21-period weighted moving average (21WMA) of the HPI, one can more easily identify the trend of the HPI. The third level of analysis is to look for divergences on both a weekly and daily basis, as weekly divergences can help identify intermediate turning points. Daily divergences, if no weekly divergences are evident, would be consistent with corrections within the intermediate trend.
NEXT: See How HPI Is Used to Analyze Crude Oil and More
|pagebreak|
Figure 1 - Click to Enlarge
The first market we will take a look at is crude oil since it has become a real focal point for the markets over the past few years, with the price of crude now often reported on the nightly news. Veteran futures traders will acknowledge that ten to fifteen years ago, if crude oil prices were discussed on the evening news, it often meant that crude was close to a turning point, but this has not been the case for the past ten years. This chart covers the last half of 2006 and all of 2007. Taking an even longer-term view of this market is difficult as the dramatic increase in the open interest makes the scaling difficult. From the start, you should note that the HPI is above the zero line for the entire period, though it did drop towards the zero line in January 2007. The HPI moved back above its declining 21WMA (point 1) one week before the downtrend in price, line A, was overcome. Crude had just undergone a sizable correction from just above $79 to the $57 per barrel level.
This brought crude back to the October 2005 lows and was a deep enough correction for many to conclude the rally in crude was over. Crude rebounded over the next six months to the $74 area before again undergoing a sharp, three-week correction. The HPI dropped below its WMA (point 2), but held well above the zero line, and four weeks later, it was back above its WMA (point 3). By September 22, the HPI had made new highs for the year and was leading prices higher. Crude oil accelerated to the upside for the rest of the year and eventually hit a high of $97.88 a barrel in January 2008. The HPI did confirm the new highs as indicated by line C, and therefore, when the HPI dropped below its WMA in mid-January (point 4), the earlier weekly confirmation of the highs suggested that this was only a correction within an uptrend. It is also interesting to note that the HPI just tested the longer-term uptrend, line B, before moving sharply back above its WMA.
Figure 2 - Click to Enlarge
The daily chart of June crude oil shows that the HPI moved back above the zero line and its WMA on January 31, 2007 (line A), two weeks before the downtrend was overcome. The first daily divergence appeared in June, and the HPI then formed a series of lower highs (line 1). The series of divergences suggested that tighter trailing stops might be a good idea. As noted on the weekly chart, the correction was sharp, but quite brief, which is typical of bull market corrections. One day after the lows on August 23, the HPI moved back above its WMA, and two days later, the HPI overcame its resistance at line 1. By September 7, the HPI had surpassed the prior highs and was acting stronger than prices.
Crude stalled in the $78 area and formed a short-term continuation pattern that was also observable on the HPI. The HPI did drop below its WMA for about eight days before resuming its uptrend. Prices then accelerated to the upside as crude reached the $93 level in early November (point 3). After a three-day pullback, crude made another new high (point 4), but the daily HPI failed to confirm this price high. Since the weekly HPI had not formed any divergences, this was consistent with a correction within the intermediate-term uptrend, and not a major top. The daily correction was confirmed when the HPI's uptrend (line 2) was broken as it declined sharply into early December. By referring to the weekly data in Figure 1, you will see that the HPI had started to rebound sharply by early 2008, suggesting that the rally in crude was not over. By February 11, the daily HPI (not shown) indicated the rally in crude oil had resumed.
NEXT: HPI Applied to Ag Commodities and US Dollar Index
|pagebreak|
Figure 3 - Click to Enlarge
For those of you who do not follow corn, it made multiple tests of the $2-per-bushel level from 1999 until late 2005, when it began its dramatic rally. From September 2006 until June 2007, it rallied over 50% and then began the period of consolidation noted on the chart above. The HPI declined sharply from the June highs and dropped back below its WMA and the zero line. By late-August 2007, the HPI was back above its WMA, and the week of September 22, the HPI moved back into positive territory. However, despite this positive action by the HPI, the correction was not over as prices retreated from the $4.20 area and tested support in the $3.70 area. The HPI dropped back to test the zero line on this correction (point 1), but held above it and the rising WMA. Though the price chart shows a series of lower highs during this period (line A), the weekly HPI formed higher highs (line B), which was a bullish indication. In early November, the resistance at line A was overcome as corn prices began their climb to the $5.50 area. The weekly HPI did form a short-term negative divergence in early 2008, but it was reversed two weeks later.
Figure 4 - Click to Enlarge
The daily chart of July 2008 corn shows the trading range discussed on the weekly chart, lines C and D. You will note that the daily HPI shows the same surge in late September as apparently significant accumulation was taking place. The daily HPI dropped back below the zero line in early October, but by the end of the month, it was back in positive territory. The price chart shows the completion of the continuation pattern on November 6 (line 1). Prices then moved sideways for over two weeks and then turned higher as the HPI resistance was overcome (point 2). The HPI continued to confirm the new highs until January 15 when corn reached a high of $5.42. Over the next five weeks, corn formed an ascending triangle (dashed lines) while the HPI diverged forming a wedge. This is the period on the weekly chart that I noted previously where the weekly formed a slight negative divergence. The HPI quickly got back in gear with prices as both the weekly and daily HPI did confirm the March 2008 highs.
Figure 5 - Click to Enlarge
The dollar index has been one of the more widely watched markets recently considering that over the past year, it has been correlated and inversely correlated with several key markets. The HPI analysis of the dollar index has been quite accurate in identifying the key turns in the dollar over the past three years. After declining for over a year, the HPI moved above its downtrend, line b, on August 8, 2008 (line 1). The HPI was already above the zero line, so the surge above the WMA was very positive. The HPI stayed above the WMA until March 21, 2009 (line 2) as the action in the HPI confirmed the negative signals from the RSI we noted at the time and set the stage for the dollar's decline into the late-2009 lows. The HPI bottomed in late October and formed a slight positive divergence at the December lows, line d. This positive divergence was confirmed the week ending December 12, 2009 as the HPI moved sharply above its WMA (line 3). The HPI has surged above the 2009 highs and is holding well above the 21WMA on the recent correction, which is bullish for the intermediate term.
Figure 6 - Click to Enlarge
Another market that got a lot of press in 2009 was sugar as it moved above 20-year resistance and reached price levels not seen since 1980. The HPI was range bound from May 2008 through April 2009 as it held above the zero line, which was consistent with positive money flow. The HPI moved through its resistance, line b, and its WMA on April 25 (line 1), a week ahead of prices. The HPI stayed above its rising WMA as sugar surged from the 15-cent level to over 30 cents by early 2010. The HPI peaked in October and then formed a negative divergence, line d, in early 2010. The drop in the HPI below its WMA and support at line e confirmed the negative divergence and indicated that an intermediate-term top was in place (line 2). The major uptrend in the HPI, line f, was broken in early March. The HPI still looks very weak, increasing the chances of a drop to the 15-cent area.
If you would like to see further examples of how the HPI works, please refer to this related article, and be sure that you understand the basic rules of open interest analysis.
