[[Start here]] → [[What works in stocks?|what works]] → [[expose to return drivers|drivers]] → [[strong stocks beat weak stocks|strong beats weak]] → smooth trends --- ![[scatter-plot.png]] Ever been in a situation where you’ve loved a stock, but waited and waited for the correction that never came? In 2023 to 2024 I watched [[2023-06-15 - Warpaint setting up for breakout|Warpaint]] for a long time as it has many of the traits of multibaggers, but I couldn’t bring myself to buy as the price trend never corrected. It turns out that shares that have the smoothest price trends have the strongest momentum effects. A recent paper titled [[Cai - Trended Momentum|Trended Momentum]][^1] delves into this idea in depth. The authors demonstrate that *shares with distinct price trends more than double the returns to a standard momentum strategy*. This has pretty profound implications both for stock pickers and systematic investors who wish to improve their investment results. More than this, I’ve found that separating the 80+ StockRank universe at Stockopedia.com into a high Trend Clarity universe enhances the returns of a 20 stock UK portfolio to more than 14.8% annualised since 2013. Let’s dive in. ### What is trend clarity? The authors of the study theorised that clear and distinct trends are more favoured by investors. They devised an online experiment involving 128 participants who were given charts to assess and rank for trend clarity. They were asked to evaluate the trends of charts from “no trend” to “very clear”. ![[smooth trends beat volatile trends.png]] If you remember statistics classes from school then you’ll have a fair memory of drawing lines of best fit across scatter plots. This is called Linear Regression, and is used extensively across many domains. As you can see below Price Path A (left) has a much better line of best fit than Price Path B (right). It has more **trend clarity**. ![[r-squared-smooth.png]] There’s a number, called the R-Squared (or R2) which defines how accurately the scatter can be defined by a line of best fit. A has an R2 above 0.8, B only 0.07. This allows a numerical calculation of the trend clarity. The authors found that using the R2 is a great proxy for finding the same kinds of charts that the survey respondents identified intuitively as having clear trends. This is a simple approach, and can be replicated in any spreadsheet application - Google Sheets or Excel. ## How useful is trend clarity? The authors separated all the stocks in the market into five buckets (quintiles) based on the 1 year historic price performance. The top bucket being the best winners, and the bottom the worst losers. They then split each bucket into five more subsets according to their Trend Clarity. It turns out that the winner portfolios with the highest trend clarity generated significant outperformance. This outperformance is independent of any other return drivers (like quality or value or size). ![[smooth trends beat volatile trends-1.png]] They also found that: * high trend clarity stocks were less likely to see longer term momentum reversals. Good news for longer holding periods. * higher trend clarity stocks had a higher likelihood of having consistently beaten analyst expectations in the year before the portfolios were formed, and the year following. * results were strongest in more mature, developed markets. Good news for UK investors. This study backs up some of the work done by Andreas Clenow in his excellent book “Stocks on the Move”.[^2] He also used the R2 of a regression to generate smooth trending portfolios of S&P 500 stocks with similarly impressive results. ## Taking it further I’ve run an adapted study on UK stocks using 6 month price strength, as many studies have shown 6 month momentum to outperform 12 month momentum. I wondered if this approach could improve the returns to the 80-100 StockRanks set. Starting from April 2013 until May 2024, which is when our archive database begins: * Starting from about 330 stocks in the 80-100 StockRank set. * Separate out the top fifth by historic 6 month price performance - about 66 stocks. * Calculate the trend clarity for each. (Linear regression → R2) * Select the top 20 by trend clarity. The results are very eye opening. A portfolio constructed in this way, and rebalanced annually has generated a capital return of 14.8% annualised. This is a 380% return over the time period - far above the results of annually rebalanced 90+ StockRank stocks over the same time period (at 214% and 10.9% annualised). (For reference, using a 1 year lookback period is not quite as strong at a 270% return.) ![[smooth trends beat volatile trends-4.png]] ## Rules of thumb for investors Ultimately, we all need some good takeaways. Let’s be clear, if you see a very smoothly rising price trend, like with Warpaint, you are unlikely to be an early buyer as you will wait for a correction. You’ll then end up buying too late. This is a mistake. We should train ourselves to buy stocks that have clear and strong price trends, as these are likely to exhibit stronger and more sustained momentum. The survey shows that we all have a very intuitive sense of what a clear trend is. You don’t need to do a linear regression - just eyeball the chart! [^1]: [[Cai - Trended Momentum]] [^2]: [[Clenow - Stocks on the Move]]