He suggested that Ted Cruz’s father played a role in the Kennedy assassination, and pledged that if he’s elected commander-in-chief, Vladimir Putin “is not going into Ukraine” when the Russian president is already there, having invaded and annexed Crimea. And the Donald swears to have seen a shocking videotape of “people taking money off the plane” that delivered $400 million in cash to Teheran, images of skulduggery that the world at large has yet to witness.

But on August 9, the day after delivering his landmark economic address in Detroit, the Republican nominee made an eminently sensible observation on the stock market. In a telephone interview on CNBC, Trump stated that “It’s all a big bubble.” Trump added that he’s dabbled in picking stocks, but, “I’m a person who doesn’t believe in it much,” and right now, “I wouldn’t do it.”

It’s highly unusual, if not unheard of, that a presidential candidate effectively advises Americans to dump stocks. And it’s unclear whether Trump bases his warning on detailed analysis, or his self-declared genius for instinctively telling great deals from ripoffs. On of that, Trump’s view that the U.S. economy is not great would certainly feed into his belief that, given that the market is near its all-time high, stocks are inflated. What’s more, Trump’s candidacy would likely get a boost by a stock drop, reinforcing his rhetoric that the Obama economy is not as good as it seems.

(For more on Trump and the stock market, read Here Are the Stocks to Buy if Trump Becomes President.)

In any case, in this instance, Trump is correct, at least according to the most reliable metric measuring whether stocks are over or under-valued.

The CAPE, or cyclically-adjusted price-earnings ratio, which was developed by Yale economist Robert Shiller, currently stands at 26.2 times average earnings, more than 60% above its more than century-old average of around 16. Anyone buying equities today is paying an extremely high, if not inflated, price.

For instance, in June of 1901, it hit 25.2, with the S&P 500 at a pre-historic 8.5. (It’s at 2,181 now.) Ten years later, the S&P stood at 9.7, a gain of less than 14%, or around 1.2% a year.

More recently, the tech investing craze of 2000 drove the CAPE to an all-time high of almost 44 in April of that year. Ten years later, the S&P had dropped 18%, from 1,461 to 1,197.

The lesson: When you’re buying at today’s CAPE, you’re getting a lot fewer dollars in earnings for every dollar you’re paying than in most periods, and history predicts that your future returns will be either low or even negative. In the CNBC interview, Trump also addressed the super-low rates that, correctly, he blames for the excessive valuations. “If rates go up,” he asserted. “You’ll see something that’s not pretty.”

On that issue, Trump’s forecast is less reliable. It’s true that the best time to buy is when the CAPE is really low, and rates are really high. Investors are weary of high interest rates because they predict lots of inflation, and companies can’t raise prices fast enough to keep up, so earnings get hit. Inflation is also a sign that the Fed and the economic braintrust are doing a poor job managing the economy, a perception that raises “risk premiums” and drives down stock prices.

But a gradual trend where rates return to normal, then stay there, would be a good thing for stocks. It would indicate that the economy is on a growth trajectory, and hence earnings are bound to follow. In the long-run, it’s simply impossible for earnings to wax at, say, 5% a year, and for 10-year rates to stay under 2%.

But on the big question of whether the market is overvalued, the Donald isn’t just making another outrageous pronouncement. Based on history and the current metrics, he’s right.