How it works

I invest for the long term. The chart doesn't tell me what to own; it only helps me decide when to buy.

Two separate questions

Is this a business worth owning? That's a judgment about the company: its moat, finances, growth, valuation and management. It's the report card on each holding's page.

Is now a good time to buy more? That's the trend rule below. I keep the two apart. A great business can be in a broken trend, and a weak one can be in a strong trend.

The trend rule

Two things decide it: where a stock sits against its 200-day average (the average closing price over roughly the last 10 months), and whether the S&P 500 is itself in a healthy trend.

Buy nowAbove the line, however far above. "Overheated" isn't a reason to wait.
OK to buyUp to 10% below a line that's still rising or flat.
Wait10%+ below the line, or below a falling line. The trigger to buy is a Friday close back above the line, not a lower price.
Buy in piecesWhen the S&P 500 is below its own falling 200-day line, spread buying out over time instead of all at once.

"Rising" or "falling" means the average moved more than 0.5% over the past month. Everything is judged on Friday closes to filter out one-day noise. Bitcoin funds use a 140-trading-day line, which matches bitcoin's 200-day line because bitcoin trades every day.

Why this rule

  • Trends tend to persist. Stocks that have been rising have tended to keep rising over the following months (Jegadeesh & Titman, 1993; Moskowitz, Ooi & Pedersen, 2012). Waiting for a dip in a healthy trend often means paying more, because the dip never comes.
  • Broken trends are where the big losses live. A simple 10-month-average rule on the S&P 500 from 1901 to 2012 roughly halved the worst drawdown with similar returns (Faber, 2007/2013). Most of the benefit came from sidestepping bear markets.
  • Most individual stocks lose to cash over their lifetimes (Bessembinder, 2018), so a stock in a broken trend deserves proof before more money goes in.
  • Earnings dates aren't a reason to wait. Stocks have tended to rise around scheduled announcements (Frazzini & Lamont, 2007).

I tested this rule on my own before using it. Past patterns are averages, not guarantees; in any single case waiting can win or lose.

Check it yourself

Every chart on this site is a live TradingView chart with the 200-day line drawn in, so you can read the rule off it directly. Charts update during market hours; my rule only counts Friday closes.