Does Market Timing Work?

As a very contrived example (to hopefully illustrate the "risk-adjusted" component of EMH while keeping the math simple), suppose the market consists of many equally sized firms and admits a strategy where each year 0.1% of firms will be uniformly randomly selected to have all their assets wiped out and distributed amongst the rest, and due to social pressures and incentives everyone uses the same strategy. Each year, 99.9% of firms will beat the market [0]. Forward-looking, 97% of funds starting in 1993 will have beaten the market that entire period (backward-looking is less meaningful because it depends on, among other things, how many new entrants to the market there are, not just their performance).

Despite the 97% success rate on a 30-year basis, this is still very plausibly a scheme you wouldn't want to participate in (your relative valuations of different outcomes might still make it desirable, but that's a separate question), but it doesn't violate EMH because of the high risk relative to the small returns being achieved.

Bringing the contrived example back to the real world a bit, that particular failure mode is common whenever a machine-learning person tries to tackle the market on their own. Even after getting over the hump of price -> bid/ask -> order book -> ... in correctly modeling what's happening, they're still prone to doing things like predicting the chance a security will go up or down and assuming that both branches have equal magnitudes. When they throw it at the real world, they find that despite low false positive and false negative rates for predicting when the price will increase, the times they were wrong were all the high-magnitude events, so they lose money on average.

Another way that potentially ties back to the real world, what exactly are the incentives for an active fund manager? When they fail, can they start a new fund? Can they distribute excess losses to a couple of years when the market also did poorly to be able to say something like "every year the S&P 500 went up, we went up more" and still attract new clients? It's not obvious to me that you'd expect behavior which would result in a low chance of beating the S&P 500 over a period of time, even if strong-form EMH holds.

[0] This assumes the "market" is static, but the details aren't meaningfully different when you instead benchmark against something like the S&P 500.

But for many retirees, they have never developed the mechanisms to continue being a productive member of society without structured work. The net result is a general drop in well being and health in their “golden” years.

Also, i made a comment where I think our relationship to work needs to change. I think that speaks to your point. I don’t think a false dichotomy between “working for the capitalist overlord” and “doing whatever I want” is what I was after.

What happens if you make this mistake in a "cash account" i.e. no margin allowed I do not know and hope not to find out by means of usual accidental carelessness.

An alternative to SWVXX is VUSB, which trades with standard ETF timing.

2. They would be much richer if they could.