You would be wrong. When looking at yearly returns they are pretty much independent.After 2 consecutive big down years [like 1973-74 and 2002-03], Monte Carlo analysis expects the same likelihood of a similar negative return in the 3rd [and 4th, 5th, 6th] year. Common sense says that when the stock market gets that cheap, more buyers emerge and returns start to turn around. At least for awhile. [The 1930s have several wild up and down years that may look random, but probably aren't.]
The reverse scenario applies too. After 4 consecutive years of greater than 20% returns in the late 1990s, is the chance of a 5th [and 6th, 7th, 8th] consecutive year of 20+% return the same as any other year? I think not.
The system doesn’t mean revert. ..
If it does it isn't showing up in the data.
A extreme example would be the Japanese market where P/E ratios hit 100 - which is a very silly number. There was no mean reversion, it just slid sideways. After moving sideways far longer than what statistical mean reversion would suggest, the bubble popped.
From https://finance.yahoo.com/news/chart-da ... 05931.html.. What you are saying is intuitive but it is not showing up in the data.
As such, it won't show up in a Monto Carlo simulation.
Maybe that 30-year downtrend from 1990-2009 is not mean reversion [a term I did not introduce to this thread].
But I don't call that a sideways move either. I call it a big, bad bear market correcting the excess rise of the 1980s.
Every single value investor believes that the market overshoots and overcorrects.
My understanding is that Monte Carlo assumes an efficient market, i.e. stocks are always priced correctly.
From https://www.edgeandodds.com/sp-500-p-e- ... -80-years/Over the past 100 years the P/E ratio has been about 15.
Sometimes the P/E ratio has been 10, sometimes 20.
During those periods is there a tendency to mean revert? No.
"The absolute lowest S&P 500 PE multiple at month-end has been registered in September 1974 at 7.0x trailing EPS.
Single-digit PEs have been recorded in four other instances: July 1982 (7.7), April 1942 (7.7), February 1948 (8.4) and June 1932 (9.4)."
I would not call the bull market after July 1982 a mean reversion.
But I do say that the SP500 with a 7.7 PE in July 1982 [which I actually remember] was undervalued.
And that Monte Carlo does not attempt to identify periods of undervalue [or overvalue].
Qualitatively, a lot of investors in Aug 1982 thought the market was undervalued and started buying.
I agree with that.Returns are driven by economic structures. Economic structures shift from time to time. These are called secular periods. During one secular period the natural mean reversion of a P/E ratio might be 10. During the next it might be 20.
..
On average these last about 10 years, but they can be as short as 3 years and as long as 20. ..
But I am terrible at predicting shifts in economic structure.
[Almost as bad as I am at market timing.] So I don't even try [to do either].
In summary, I don't think Monte Carlo is wrong or bad or even inaccurate.
I just think it has flaws, like almost everything.
Statistics: Posted by doobiedoo — Sat May 25, 2024 9:45 pm — Replies 26 — Views 2967