high low

¹ u.s. bonds < 9z-fbb «>

§ ii il It ! I SS S i I

quite different. Tliey are also remarkably close to the pattem seen in 1Q76 cattle, except that the end of day is not as pronounced in the recent period.

In general, 25**o to 40**o of all highs and lows occur in the first 15 minutes of trading. Afterward, there seems to be an equal chance of a high or low occurring throughout the day. If we look closer, we can see a slightly larger increase just after 12:00 onmost charts,

FIGURE 15-5 (Continued)

and an additional increase in the ftequencj of highs and lows near the close. Eurodollars show the greatest chance of a high or a low occurring at the open, while coffee has the same 40** chance over the first 1 hour, 45 minutes of trading. The S&P fiitures makes 25**o of its highs and lows in the last 1 hour, 15 minutes of trading, with 10<*o after the NewYork Stock Exchange closes at 300 P.m

Trading Decisions

For maikets such as the Deutschemark, which has a steady chance of posting a new high or low anjiinie after the first 15 minutes (remember that this inter\al includes any reaction to U.S. economic data that is released 10 minutes after the open), trading could be approached by deciding, after 15 minutes, whether the market has already made a high or a low for the day. For example, if an economic report indicates a weak dollar, and the price of the D-mark jumps 50 ticks, we could reasonably assume that we have seen the lows of the day. If prices make a new high during the next 15 minutes, we are even more likely to have seen the lows in the first 15 minutes. It is possible that we have seen both the lows and the highs, bul it A-as the lows only that occurred at the most common time. If we assume that the high can occur anjiime during the day, then bujing sooner wiH give more opportunity to profit by the end of the day.

You can increase the confidence in your decision by waiting about 2 hours into the trading day, when most maikets have a50**o chance of seeing a high or alow already estabfished. While this may make it easier to identify the correct extreme, it reduces the opportunity for profit by the end of the trading day. A projection of volatility, which can be reasonably constant for many markets, could help decide whether there is enough opportunitv to enter a trade.

OPENING GAPS AND INTRADAY PATTERNS

The daily price fiuctuations and intraday pattems of a maiket may be related to the strength or

weakness of the dauy opening. A large gap opening implies greater volatility for the day, positions prices for a greater reaction, and may show more consistencj in the subsequent intraday pattems than a maiket thai opens nearly imchanged. A gap analjsis was performed to study the combination of pattems that could be traded profitably. A strong open followed by a stronger close would allow a trader to buy on a pullback after the open, while a consistently weak close after a strong open would be an opportunity to set a short position. Specific entry tuning can be improved by using the pattems discussed in the preus section.

The gap analjses for the six diverse maikets shown in Table 15-6 are separated into two parts:

1 . The position of the closing price following a gap opening.

a. The close continued (extended the move) in the direction of the gap

b. The close was between the gap opening and the prior close.

c. The close reversed the direction of the opening gap (that is, if the open was higher, the close was lower).

2. The tjpe of trading range pattem that occurred.

a. Prices crossed the prior closing price at least once during the trading session.

b. Prices reversed ftom the open but did not cross the prior close.

c. Prices continued in the direction of the open and never reversed.

In addition, a record is shown of the percentage of times that pnce continued on the following open in the same direction as the current gap.

Using the Deutschemaik as an example, we see that the gaps ranged ftom +1.00 to -1.00 in IMM notation. The values in each of the columns, except for the far right column, are the percentage of occurrences based on the total cases, shown on the right. The numbei of cases show a tjiiical price distribution, clustered near the smaller gaps, with an extended tail at the extremes. All gaps greater than +1.00 were included in the top row: all gaps less than -1.00 are included in the bottom row. The interesting pattems in the Deutschemark results are:

TABLE 15-6Gap Analyses, Januarj 1986-Februaij 1996*