For Total US Market - try https://fred.stlouisfed.org/series/WILL5000IND or https://fred.stlouisfed.org/series/WILL5000INDFC ; I believe one of these is the TR data (with dividends reinvested); https://fred.stlouisfed.org/series/WILL5000PR is the price only (i.e. no dividends) data. The Wilshire 5000 TR data goes back to 1970 or so, I believe; I can get access to CRSP US total maret data as well but the version I have only goes back to 1989 rather than 1926 (which is where the full version starts at).StrategyDriven wrote: ↑Tue Apr 06, 2021 5:51 pmRegarding back testing, if I can get Total US Market, or even just S&P 500 monthly total return data AND a T-Bill/CASHX type monthly total return then I can tell you which months the strategies would have been out of equities. Testing on IWB and VTI or SPY does make small differences, I use IWB for my strategies which use Russell Components, and VTI for Global Navigator which has no Russell components. But they're all extremely similar other than maybe a month here or there which is in vs out.D1984 wrote: ↑Tue Apr 06, 2021 5:28 pmGood to know.StrategyDriven wrote: ↑Mon Apr 05, 2021 6:02 pmbtw - I just recalled hearing Meb Faber talk about Black Monday, and that if your timing strategy was faster than a 200 day, you got out before the tumble, which means that my strategies would have all been out before that day, they move faster than 200 by quite a lot. This past year they got out of the markets at the end of February. The Pro is getting in and out a bit quicker, the Con is sometimes whipsaw.
I wonder what would happen to strategies like these in a sudden no-warning "thunderbolt out of the blue" crash like in May 1940, March 1939, or March 1938 (although for that last one equities might've been in enough of a downtrend already that the system would've been in bonds by then anyhow). I know the Russell indexes don't go back that far but there might be a good enough substitute....see, both Siamond on the Bogleheads board (they used data from Ken French at Dartmouth) and Tyler--of this very board and of PortfolioCharts fame--(he used CRSP data IIRC) have computed monthly TR indexes for all nine sectors of the style box (i.e. LCG, LCB, LCV, MCG, MCB, MCV, SCG, SCB, SCV) going back to the mid-1920s. These should be fairly accurate proxies for their respective equivalent Russell indices (i.e. LCB for Russell 1000, MCG for Russell Midcap Growth, SCV, for Russell 2000 Value, etc) since Russell doesn't have earnings quality and positive-for-at-least-four-quarters earnings requirements for entry into its indices like S&P does (for instance, recall that Tesla was only allowed into the S&P 500 when it could show four consecutive quarters of positive earnings); Russell just requires the companies to be of the proper size for the index category (for instance, large-cap for the Russell 1000) and being value or growth stocks for the value or growth sub-index versions of its indices.
If I could get you this data as well could you please run the backtests on these back to 1926 or 1927 (if you need monthly LTT or ITT data back that far Siamond at Bogleheads should have that as well)?
As for the S&P 500 TR the issue is that there wasn't actually a "true" S&P 500 TR until March 1957; if you look at the historical "S&P 500 TR" data there are at LEAST six or seven different version running around (the Cowles one from 1871 to the late 1930s--this is the earliest one and the one Shiller uses for his S&P and CAPE dataset portion from 1871 to 1936 or so; the S&P 233 Composite from 1923, the S&P 90 Composite starting in 1926, the one starting in 1936, the one starting in 1970, the one starting in 1976 when financial stocks were first included, the one starting in 1988 or 1989 when daily TR data with dividends was provided, the "CRSP version of the S&P 500" using all the stocks on the NYSE, and various "S&P" composites with anywhere from 198 stocks--in 1918--to 480 stocks--in the mid-1950s--and that in some cases actually predate the 1941 merger between Standard Statistics and Poor's Publishing that created the modern-day Standard and Poor's Co).
For more background on all these (including with some useful links giving monthly and/or daily TR returns for the 1936, 1970, and late 1980s version of the S&P 500) see the following websites:
https://www.cftech.com/the-brainbank-ar ... poor-s-500
Finally, if you want monthly TRs for the S&P 500 (no idea which of the above versions or whether indeed it is a concatenated splice of several of the above indexes) the 1928-72 data is at https://static.seekingalpha.com/uploads ... origin.png ; the 1973-2017 data is at https://static.seekingalpha.com/uploads ... origin.png
As far as T-bill/cash total return data goes I'm still working on it; I should be able to provide a series back to at least the late 1930s or early 1940s.