Valuation: Random Musings on Mean Reversion and Capitalizing R&D

So the way the Valuation course works is that all materials are available online – lecture videos and slides – and it’s an entirely self-driven process. Every couple weeks, however, we are able to get an hour of one-on-one with Professor Damodaran and ask him any questions that might be on our minds regarding the course. This week, there were 2 interesting points that caught my attention: the notion of mean reversion in the markets and why you should capitalize R&D expenses.

Mean Reversion

In the back of my mind, I continue to wonder whether the stock market returns the US experienced over the past 100 years or so was normal or an anomaly. I tend to lean slightly towards anomaly. Why? Because the world witnessed two world wars, with the second one basically destroying the economy of all other advanced economies except the US.

The US after WWII was the last country standing. The US was the only country that could supply the rest of the world with its exports to rebuild the world after WWII. Therefore, of course the stock market returns experience by the US was rather special in the 20th century, especially after WWII.

I was reading this article on the economic Golden Age in the West from the late-1940s to the early-1970s which reminded me about what Damodaran said during the recent Q&A we got with him online.

He said that around 90% of active investors invest money based on the notion of mean reversion. I’m going to be honest, I tend to make this sort of lazy assumption too when I see some stocks get heavily sold off over one day and take a small stake. Damodaran says that mean reversion requires a stable system. The US in the past 100 years has experience the most stable economic system in history. From 1926 to 2000, mean reversion worked really, really well because the system in the US was inherently stable.

Concepts such as buying low P/E, a PEG ratio of 1 or under, low Price-to-Book, etc were all based on having a stable system in place over the last 80 years or so of US history. However, Damodaran believes after 2000, the system has changed – it really has been “different this time” since 2000. He believes that 2008 was a break in the system and that we are in a new era where what worked in the past might not work so well anymore going into the future.

Someone asked him what he thought about the Shiller P/E and he said that the Shiller P/E was a lazy way of thinking: saying that the Shiller P/E is above historical averages and thus the market is overpriced is a lazy way of thinking according to Damodaran. Since he thinks the system is not the same since 2008, how could you rely on past data to forecast the future of a different economic/market system?

I found it an interesting topic to chew over. I am not yet confident as to what I think in any strong way yet. This is a topic I will continue to chew over.

Capitalizing R&D

Currently, R&D is expensed. This means that net income is perhaps depressed. Apparently, accounting rules will be changing in a fews years that will see these items being capitalized rather than expensed.

Think of the purchase of a new factory for your business. While you have to pay for the factory in cash today, you get to capitalize the factory over its useful life – in this case, let’s say you capitalize it over 20 years. If the factory cost a million dollars, capitalizing this expense means you get to record the “cost” of this expense as $50,000 every year for 20 years, effectively spreading out the cost of the expense over its useful life.

This is done because the factory will be a productive asset that generates revenue and profit for your company for the next 20 years. It wouldn’t be entirely accurate to deduct a million dollars in the first year on your income statement seeing as the million dollars was used to purchase a productive asset that is expected to provide use over the next 20 years. Therefore, you get to depreciate the cost.

With R&D, this is currently not the case: R&D is expensed in the year it is incurred, which means that the total spent on R&D is subtracted from the income statement right away when it occur.

But if you think about it, for something like a tech or pharmaceutical company, money spent on R&D is going to far outstrip any money spent buying land, buildings, factories, etc. R&D is essential to these companies for producing future revenue and profit. Therefore, it would make more logical sense that you would capitalize and depreciate/amortize R&D expenses over its useful lifetime.

Damodaran stated that tech companies have big R&D spending and these are all expensed rather than capitalized, effectively depressing tech company earnings. Additionally, tech companies now make up a large portion of the S&P 500. In the past the S&P 500 was made up of more traditional operating companies that capitalized their CAPEX such as land, equipment, buildings, factories, etc. Therefore, todays seemingly “high” P/E for the S&P 500 might not actually be that high if you re-adjusted R&D spending by the large component of tech companies as capitalized expenses.

If found this another interesting opinion presented by Damodaran.

Recommended Readings