Differentiating Between Trending vs. Mean Reverting Data

Thursday, December 02, 2010

Share
Derek Hernquist had a great post out a few weeks back titled "How I Use Mean Reversion" in which he discussed the stark difference between two types of data: trending and mean reverting. While these ideas are not rocket science, it is crucial for success in markets to understand how these two types of data differ.

What are mean reverting data?

Hernquist points out some typical mean reverting data: sentiment, valuation, profit margins, asset class popularity. I like to view many of these in terms of a swinging pendulum. Sentiment, in particular, swings back and forth between extremes around a static mean. Over long horizons humans really don't change. We are greedy at times and fearful at others and these sentiment changes will effect valuations on assets.

Financial excess has never been and will never be stopped. Some level of government intervention is helpful in muting the extremes, but regulation cannot fix this eternal cycle. Financial markets will always find a way expand to irrational levels and contract to similarly irrational levels. Humans, given a recency effect, tend to believe good times will continue forever, likewise bad times.

In the same vein profit margins drive asset class popularity. Yet profit margins will mean revert over time. Economic competition keeps industries in line with each other. High margins induce new firms to enter and squeeze margins while low margins incent firms to exit.

Hernquist's best observation from his post is that "there is a general tendency for news flow to continue in one direction, but an oscillating pattern in how market participants respond to that news flow." We see this everyday in the stock market and turning points can often be foreseen by gauging reaction to headlines. When prices stop declining on bad news, often a turn is afoot.

What is trending data?

Examples of trending data are highlighted as: sales, earnings, government spending, stock prices. While I think it could be argued in the very long term, sales and earnings data for individual companies do exhibit some form of mean reversion due to the competition factor I discussed above, trending does occur at least on a 5 year horizon, plenty of time for investors and traders. In the case of the market or economy as a whole, the data are most certainly trending in nature.

If you have any faith in human innovation and ingenuity, you'll recognize that stock prices will not fall in perpetuity at any point. Stock prices, for all their volatility and randomness, follow earnings over the long-term and tend to rise as economies expand and humans create value to sell to each other. The chart of the Dow Jones Industrial Average for the last century tells the story better than anything else can.


Does this mean anything for the active trader?

In my view, shorting is a useful strategy during anomalous times like the Panic of 2008 and for very select companies about which you know a great deal. The market, though, is built to go higher. Very simply, we drop failing companies from the index. Economically, companies go bankrupt. Removing failures from an index and replacing them with growing firms helps push prices higher. There's just no reason to believe in lower prices over the long-term.

When looking at individual companies, it's important to see that momentum and trends tend to persist. Shorting flying stocks is often a losing, and almost more importantly, extremely frustrating game played in vain. Innovative companies expanding their sales and earnings can have dramatic rises in price and not be in a bubble or nearing some impending collapse. The persistent shorts in names like Apple (AAPL), Netflix (NFLX), Salesforce.com (CRM) or Baidu (BIDU) fail to recognize the trending nature of earnings and stock prices. Quick trades, scalps and the like are always possible in any moving price but often sticking with trends will be the most successful strategy in the long run.

It all comes back to the old statement: don't fight the trend. Stock prices trend along with the sales of companies and there's just no reason to doubt it. As I talked about yesterday, we are in a recovery with growing GDP and corporate profits and these are trending data. Separate this data from mean reverting indicators and the trend will become clearer and easier to stick with.


Brandon R. Rowley
"Chance favors the prepared mind."


*DISCLOSURE: Long AAPL.
blog comments powered by Disqus