# Beware of Buying the Dips, 5/18

### Recording of Monday night's discussion of financial conditions

Monday night, for paid subscribers, I discussed financial conditions with Matt Gelfand and Alan Marcus. The recording is here. Cameo appearances by Lee Bressler, Mike Lorenz, and Tom Grey.

Because of technical issues, I cut out the first few minutes. With no introduction, Matt starts talking about his doubts about “buying the dips.” Today there was quite a dip, and if you’re thinking of buying, you might want to watch the video first.

The discussion left me so pessimistic that I sold some stock yesterday. I wish I had sold more.

Comments welcome, but be gentle (but fair) with your style criticisms. We’re not professionals at this, and it takes us a while to articulate our thinking.

Re: risk increasing rather than decreasing when investing in the stock market over time: this depends on what you are comparing to.

Starting with the toy example of betting $1 on one roulette wheel spin, vs. betting $1 each time on 1000 roulette wheel spins: yes, the absolute dispersion of results is larger in the second case rather than the first. More precisely, under the standard assumption that the spins are independent, the dispersion is sqrt(1000) = ~31.6 times as large. However, if you replace the first case with betting $1000 on one roulette wheel spin (so that we're equating the total value bet across time between the two cases, instead of the value of each individual bet), it's now the first case that has the larger dispersion (by the same factor of sqrt(1000), it turns out).

Similarly, if you invest $1000 in a random stock at a random time and sell after 4 years (~1000 trading days), the absolute dispersion of results is greater than it would be if you only hold for 1 trading day. But the dispersion increases sublinearly as a function of time, you can further reduce it by e.g. investing in a broader stock market index rather than a single stock, ...and perhaps most critically, many large expenses (housing, college tuition...) have been rising at a rate closer to the average stock market return rate than the T-bill rate over the last several decades.

Because of this last point, "risk-free" investment is more like "guaranteed-loss" investment. If you don't want to lose purchasing power over time, you basically have to take a certain amount of risk, and it is worthwhile to determine lower-dispersion ways to do so even if no solution has quite as low dispersion as you'd like.

For an investor with a long-term perspective, I have some trouble understanding the concerns here. Generally, markets are subject to periodic bouts of euphoria and panic, and those sentiment swings always have rational sounding bases for them at the time. But at the end of the day, despite the gyrations, the stock market has generated substantial real returns over time.

Ultimately, it seems to me, the question is whether the asset in question is reasonably priced at a given moment in time. And if you apply fairly conservative cash flow assumptions and an appropriate discount rate (and I realize there is some play in the joints here) to many U.S. high quality companies — say an Apple or a Microsoft, for example — they seem to be trading at prices (as of today) well-below their actual values.