FINDING THE TREND
One question that comes up often is how to draw the original trendlines. The simple answer is that if you look for confirmation, then you will know that these lines resulted in accurate FLIRs, and that’s all that’s necessary.
The number of sets of trendlines is up to you. Often, confirmation will require you to create an extra set or two. What’s interesting is that the extension of any trendlines will still provide a readable framework for FLIR, so even old lines extended into the future will provide forecasts.
Reverse confirmation may seem easier and more confirming. You simply find a top or bottom, then measure the distance at that point between trendline axes, create a horizontal line equal to that length, then drag it around to see if it matches the distance between any pair of highs or lows, or between a high and a low.
During confirmation or beta testing, if you notice that FLIR is consistently off by exactly the same amount of time, you still have a reporting setup. Just adjust your date or time
In FLIR, momentum change with continuation can be deadly. But it need not be feared, as long as you set correct stops or are quick on the trigger. Knowing when it’s happening is pretty easy.
Ideally, the period immediately after a true FLIR should exhibit a higher high than the FLIR (in a reversal to the upside), and a lower low than the FLIR (in a reversal to the downside). If it should be reversing to the upside but instead breaks lower than the low of the FLIR, or if it should be reversing to the downside but instead breaks higher than the high of the FLIR, it’s time to get out because you are almost certainly looking at momentum change with continuation, and it’s going to get volatile quickly. If using stops, you want to set those just above or just below the high or low of the FLIR (see “Stop. Do not pass go”).
It doesn’t matter which tops or bottoms you measure. Any combination should generate a FLIR if you’ve had a couple of successful confirmations. In fact, you need to use different spans so you can generate several forecast points.
On this same subject, not all FLIRs are of equal value. That is, some will only identify mini trends of just two periods — in other words, whipsaws. For safety’s sake, that probably leaves FLIR most often as a short-term scalping tool. So it’s wise policy to exit your FLIR trade the same day or period as you entered. Don’t let greed get the better of you. If the issue shows clear signs of continuing to run in your direction, you can always re-enter.
That said, FLIR does have a distinct propensity for forecasting macro-reversals of considerable duration, even when the forecasted FLIR is technically a failure (the FLIR exhibits neither a higher high nor lower low than the preceding period, nor is it within the plus or minus one-day rule). In “Be it ever so wrong” (left), there’s a clear reversal almost immediately, but sometimes charts dither briefly, with a few small whipsaws, before finally fulfilling the FLIR forecast.
Caution is also advised with weekly and monthly charts. The problem is the slippage issue mentioned earlier. To be off by one period on a daily or intraday chart is usually no big deal because price is relatively stable across one or two such periods. But to be off by one period in a weekly or monthly chart can be unforgiving due to the vast strides that price can make against you in that time. So unless your confirmation shows perfect timing (using the same trendline pair as used in your forecast), you may wish to avoid weekly and monthly charts until you obtain the necessary precision.
FLIR’s accuracy can lead to trader over-confidence, even to the point of trying to trade the FLIR itself. Say you’re looking at a long downtrend when FLIR occurs. You naturally assume that FLIR will mark a reversal to the upside, right? And that may happen, or not. So sometime during the period you pick what you think is a nice low spot at support and go long. But suddenly support is broken and you realize you’ve just caught the Momentum Change Express to Doomsville. Or, let’s say you pick that nice low spot and go long, price duly reverses and puts you in the green but then a funny thing happens: It goes so far green that it makes the FLIR itself dubious (see the fourth example in “FLIR or no FLIR?”). Do you hang on, or take the money and run? You should probably take the money and run. The lesson here is that trading on the FLIR itself is tricky. Wait for the period after the FLIR to enter. Greed can be fatal.
To be sure, using FLIR to make money is a great idea, but FLIR may have a future well beyond this. Earthquake and volcano prediction come to mind, as well as medical applications, fluid dynamics, and environments where strain forecasting may be important to cost saving or life saving, such as in heavy machinery or aircraft components. It will be interesting to see if and how FLIR migrates from investing to other areas.
Dennis Hudson created FLIR. He honed his quant skills at General Electric Co. in the 1960s, in forecasting particle size distributions in diamonds manufacturing. He’s an active options trader and principal of OrangeQuant.com, where he can be reached.