David Keller's Three Timeframes of Technical Analysis

The first thing David did was set the stage by making a comparison between Golf and Investing. While I've never golfed, nor have I watched it, his analogy made a lot of sense. He asked our group what type of club we would use in a variety of situations; driving, chipping, putting, etc. and those of us who knew golf answered accordingly. He then spoke about an instance where his father used an 8-iron off the tee and hit a perfect shot toward the green, but only to see if fall short due to the headwind and the upward slope the hole was set on. Under normal conditions it would've been the right club, but in technical analysis and golf alike, different environments call for different tools.

Dave would use this concept of knowing the right tool for the right environment throughout his presentation, but starting with the basics. Using his hand-drawn chart (fluctuations courtesy of his United Flight to NY) he explained that there are three timeframes; short-term (blue), medium-term (orange), and long-term (black).

Recognizing that most market participants generally have different meanings for those timeframes based on their particular approach, he shared with us how he defines them and what their main uses are in his process using the great illustration below.

He uses the short-term, defined as a couple days to a couple weeks, as a tactical tool because prices tend to mean revert over this time period.

He uses the medium-term, defined  as a couple of months to a couple of years, primarily for trend following.

He uses the long-term, defined as many years, primarily for market context as prices once tend to revert back toward their fundamental value.

On the bottom of the slide we see that the value of fundamental analysis increases over time as prices are more likely to attain their intrinsic values over the long-term. While the tactical contributions from Technicians may be more limited over these longer time periods, what doesn't change is our ability to provide the behavioral perspective on long-term trends that would otherwise go unnoticed without analyzing price and derivatives of price. Since history has shown us prices can diverge from their fundamental value for extended periods of time, even fundamental investors need to be stay aware of the behavior of the market and its participants to manage risk. Or as John Maynard Keynes would say "The market can remain irrational longer than you can remain solvent."

The next topic we spoke about was the the three types of value we obtain when we use technical tools. Their purposes can be bucketed into three main categories: predictive (used to predict), descriptive (used to describe), and comparative (used to compare). Keller notes that tools can fall into more than one of these categories as different tools serve different purposes based on their timeframe.

For example, a 5-period RSI on a daily timeframe can be used to help improve entries and exits by identifying divergences in momentum relative to price. In this case, RSI is being used as a predictive tool.

A 14-period RSI on a daily timeframe, while also predictive if used to identify divergences for the purposes of predicting reversals in price, is more descriptive in nature. Using RSI over this longer period helps identify whether momentum is supporting the trend in prices.

Momentum is said to be in a bullish range when it's hitting overbought conditions (above 70) and does not hit oversold conditions (below 30) during corrective periods. This is the type of confirming action we want to see in uptrends. On the other hand, momentum is said to be in a bearish range when it's consistently hitting oversold conditions (below 30) and does not hit overbought conditions (above 70) during corrective periods. This is the type of confirming action we want to see in downtrends.

The last category we haven't touched on yet is comparative. There are many different tools that fit into this category, but one we discussed during Dave's presentation was the Relative Rotation Graph. Essentially this graph charts the performance of the securities you select relative to a benchmark of your choosing. The example below is a weekly relative rotation graph of the S&P 500 sectors relative to the S&P 500. Each line has 5 tails and each data point along that tail represents one week of performance. As performance changes, you'll see the sectors start to rotate around the graph's four quadrants of leading, weakening, lagging, and improving. This game is all about relative performance, especially if you have career risk, which is why comparative charts are essential in determining your opinion on one security or portfolio versus another.

The last topic we spoke about was long-term charts. This was an interesting topic because while long-term charts are good for historical context, they're typically not useful for the purposes of generating signals. In the US we are lucky to have robust financial markets with a lot of historical market data, however, even we only have a handful of decades of history before the data starts to get wonky. In less robust financial markets like India, there is even less good data available.
The example used to illustrate this was the 10-year treasury yield. We've been in a bond bull market for several decades, but now we're starting to see a change in behavior with rates now going sideways rather than continuing lower. What does this mean for rates going forward? Nobody knows. We have no historical precedence to base our conclusion on, so anyone claiming they know where (and how quickly) rates are going with any sort of conviction is not being truthful to themselves.

The Bottom Line: David Keller is a wicked smart guy that we can all learn from and his presentation on timeframes came at just the right time. There's been a lot of noise in the short-term regarding the broader market, but I have a feeling this presentation helped re-frame some mindsets in the audience, mine included.

If you enjoyed this summary I'd encourage you to connect with Dave directly via his Blog, Twitter, or LinkedIn and check out his firm Sierra Alpha Research. He'd love to hear from you.

Also, if you're in the New York area we'd also love to see you at the next NY Chapter meeting of the CMT Association.

Thanks for reading!

Allstarcharts Team