Time versus Volume bars

As discussed earlier, true volume, the number of contracts traded, is not available for futures on a live broadcast basis. Nevertheless, we can use tick volume as a proxy.

A number of charting packages such as Omega Research TradeStation and Aspen Graphics provide tick volume bars. Tick volume bars simply increment each bar by the number of ticks as opposed to the number of minutes. With regard to a tick bar, the open is the price of the first transaction that occurred in that bars set of ticks, the high is the highest priced transaction in that set, the low is the lowest priced transaction, and the close is the last transaction of the set.

Tick volume bars are superior to time bars in most cases because they form more quickly during periods of high activity and more slowly in periods of quite activity. This produces visually cleaner signals.

We can compare the 15-minute March 1997 Wheat chart (Figure 12.11) with the same chart shown using 28-tick bars (Figure 12.12). Looking at the KAR lines, we see that on average the risks in both scenarios are identical, that is, right around 3.25 cents per bushel. Thus, we have a similar global risk environment. Again we emphasize that the 3.25 cents is the average risk.

WH7-15 rain

HI"l"T

I It*

4,V,rt

398A3 I

395A0 J

•391A1 I 387 2 I •383A3 I

*KAR 3.25 5.48 8.17 11.30

1 8 9 4Kb l/13

Although global risk is similar, the tick chart offers more information. Compare point a on both charts. At point a on the tick chart, we see a hanging man followed by four down candles. On the time chart, we simply have three down candles. This does not give us the same amount of warning or chance to enter on the short side that the more elongated price activity on the tick chart provides.

Similarly, at point b on the tick chart, we have a Harami line and star followed by a bearish engulfing line, very similar to a modified evening star formation. On the

Figure 12.12 Twenty-eight tick WH7 wheat-average risk at 3.22 cents.

WH7-28 Tick Bars

"Ill-

398A3 I

•395 0 I 391A1 J 387 2 j •383A3 1

*KAR 3.22 5.32 7.84 10.78

1 i/os Tre i/io

12:24 1/13

Figure 12.11 Fifteen-minute WH7 wheat-average risk at 3.25 cents.

15-minute chart, we simply get a closing point reversal. Thus, we get a high degree of warning that a reversal may be coming with the Harami line and star on the tick chart. On the 15-minute chart, we get no warning of a reversal until after the closing point reversal is complete with a lower close, and the market already headed down.

Quantitatively, tick charts offer less risk. When comparing Figures 12.11 and 12.12, we can see as we scan across the average risk line that the first standard deviation over the mean is higher on the 15-minute chart. At the second and third stop levels, we see the same pattern of slightly higher risk to where we have almost a half-cent more risk on the 15-minute chart than on the tick chart. This pattern is consistent not only on these charts but on all markets in general. Our research has shown that, on average, we can expect a reduction in risk by something along the order of 15 percent to 20 percent using tick charts. While this may not be significant to a trader on an individual trade, a 15 percent to 20 percent savings in losses consistently throughout years of trading will add up and become significant. Thus, when trading intraday bars, certainly of a third of a day or less, tick bars provide a superior trading vehicle.

When setting tick volume bars, we recommend using a Fibonacci number (such as 8 ticks per bar, or 34 ticks per bar) or the average of two consecutive Fibonacci numbers (such as 28 ticks per bar) to meet risk criteria. For a time-based chart, we recommend that on average the number of bars per day be equivalent to a Fibonacci number as well.

multiple Time Frame trading

Scaling Up in Time

Since risk is related to the time frame or bar length we trade, then it stands to reason that the lower the bar length, the less the risk. Even if we are willing to trade a longer time frame, the risks associated with that time frame may be more than it is necessary for us to bear. Also, the longer the time frame, the less frequently we will see trends. For example, if we are trading a 5-minute chart, we might get a "trend" two or three times a week. If we are trading a daily chart, we might only get a trend once or twice a year that lasts sufficiently long for us to trade it.

"Scaling up in time" means to increase the number of bars we would see in a days worth of trading. By scaling up in time, we see dramatic risk reduction and an increase in our chances of catching a trend. So we advise initiating trades on the shortest time frame that is practical, for example, 15-minute bars. Assuming the trade is not stopped out and then confirmed in a larger time frame, we move to the larger time frame, for example, 45-minute bars. If we are not stopped out in this intermediate time frame and get a confirming signal in even a larger time frame (e.g., daily bars), then we can move to the larger time frame and manage the trade from there.

Three charts are shown in Figure 12.13. This is again the March Wheat chart, showing a trade in November 1996. We can see on the 15-minute chart on November