Seems like quite a few people making this argument recently. It feels more coordinated than the raw numbers can justify. It's not like this level of market concentration happened overnight. But I guess bears just come out whenever a new ATH is reached.
> Now think back to how you felt during the heat of the sharp decline for the S&P 500 through April 8. Did you wish you had a more diversified — or less risky — portfolio?
Probably many people did, and took off risk at the exact wrong time, missing the recovery.
> Mark Hulbert spelled this out in an article about investment-portfolio diversification last week: “In the past, when the market was overvalued as it is now, a 60/40 portfolio almost always beat the S&P 500 over the subsequent decade.”
This is the crux of the argument made throughout the article. The problem is the premise that the market must be "overvalued" today compared to historical levels, because of metrics like P/E ratios compared to historical levels. But this ignores that the fundamentals are different. Overwhelmingly, the successful companies today are ones that have been reinvesting profits or doing stock buybacks rather than paying out a dividend. (This is, per Buffett's ideas, something that investors should be agnostic to — total return is total return — but this fails due differences in tax treatment between being paid 1% of the stock's current value as a "dividend" and selling 1% of your holding and realizing a "capital gain". So many companies now prefer to not pay a dividend and expect investors to make their own dividends through sale of appreciated stock.)
The old rules are based on a model of stocks that tread water while paying a hefty dividend, and expecting that dividend to exceed the risk-free rate of return plus a risk premium, averaged out of the long term. But it turns out that maintaining a functioning economy post-gold-standard requires considerable fluctuation over time in the risk-free rate of return. Meanwhile, of course if your investment instrument is designed with the expectation of appreciating over time representing the accumulated wealth of the company, the price is going to reach higher multiples of earnings. Because that wealth is a function of the integral of earnings over time.
People making this argument are not saying "this time it's different". They're pointing at decades-long trends and shifts, and adding them up to say "nowadays things are different in general". There isn't a "this time", because the thresholds for declaring a "time" are entirely arbitrary.
zahlman•1h ago
> Now think back to how you felt during the heat of the sharp decline for the S&P 500 through April 8. Did you wish you had a more diversified — or less risky — portfolio?
Probably many people did, and took off risk at the exact wrong time, missing the recovery.
> Mark Hulbert spelled this out in an article about investment-portfolio diversification last week: “In the past, when the market was overvalued as it is now, a 60/40 portfolio almost always beat the S&P 500 over the subsequent decade.”
This is the crux of the argument made throughout the article. The problem is the premise that the market must be "overvalued" today compared to historical levels, because of metrics like P/E ratios compared to historical levels. But this ignores that the fundamentals are different. Overwhelmingly, the successful companies today are ones that have been reinvesting profits or doing stock buybacks rather than paying out a dividend. (This is, per Buffett's ideas, something that investors should be agnostic to — total return is total return — but this fails due differences in tax treatment between being paid 1% of the stock's current value as a "dividend" and selling 1% of your holding and realizing a "capital gain". So many companies now prefer to not pay a dividend and expect investors to make their own dividends through sale of appreciated stock.)
The old rules are based on a model of stocks that tread water while paying a hefty dividend, and expecting that dividend to exceed the risk-free rate of return plus a risk premium, averaged out of the long term. But it turns out that maintaining a functioning economy post-gold-standard requires considerable fluctuation over time in the risk-free rate of return. Meanwhile, of course if your investment instrument is designed with the expectation of appreciating over time representing the accumulated wealth of the company, the price is going to reach higher multiples of earnings. Because that wealth is a function of the integral of earnings over time.
People making this argument are not saying "this time it's different". They're pointing at decades-long trends and shifts, and adding them up to say "nowadays things are different in general". There isn't a "this time", because the thresholds for declaring a "time" are entirely arbitrary.