Hi Tom
Historic (typically) 30 year SWR measures identifies the individual worst case withdrawal rate period. That individual worst case often follows large/fast gain up-run, euphoria - over-extending share prices and then subsequent realism drop. Think Wall Street Crash 1930's - that followed the "Roaring 20's" when all, including shoe shine boys, were 'experts' at making money from stocks, some/many even massively leveraged stocks (borrowed to buy shares). Japan's 1970/1980's is another example, the rise of Yamaha, Sony ...etc. to global household names, and the subsequent 1990's corrective declines. Dot com bubble 1999 peak, perhaps a AI bubble next. Any method that has you reduce being heavily in at euphoric highs is inclined to improve the SWR outcome. More often the average actual SWR supported is considerably higher than the 4% guideline SWR commonly suggested (for all-stock 3.5%, but many opt for 4% on the assumption that they wont be in the worst case 5%)
I applied another AIM measure, standard AIM (10% SAFE, 5% MTS) but where AIM-CASH is set to zero if/when it transitions to negative (so first AIM Sell after a deep dive has some positive AIM CASH), I also applied a 2% Portfolio Control acceleration for years since 1985 to account for the transition towards stocks retaining more of earnings, paying out less in dividends. I further limited resulted AIM CASH, that otherwise ranged 0% to around 66% to >=20% <=50% i.e. between Robert Lichello's 80/20 AIM-HI and standard AIM 50/50 choices (if <20% then indicate 20%, if >50% indicate 50%). From that AIM that dates back to 1915 I extracted out the 1982 years for which vwave/iwave data is available. As that AIM averaged 36% cash I adjusted both vwave and iwave to also average the same, average cash equalization. For vWave (that is the index form, individual stock form / 1.5) that involved adding +8% to each yearly value. Ditto iwave but where +11% was added. So all three averaged 36% cash across 1982 to 2023 years. Aligning actual portfolio value once/year to those indicated %CASH amounts and it was a close call between each of the three, ranking best to worst ... iWave, vWave, AIM. However the differences between all three might be considered insignificant, in the scale of things too small to be considered potentially nothing other than 'noise' differences.
Fundamentally by providing liquidity (reducing-high) to the greedy, and to the scared (adding-low), mostly its a case of reducing the pain from periodic sizable declines from over-extensions that seems to be the primary driver factor. Consider another case of a target 67/33 stock/gold asset allocation. What might drive 67 stock value to halve to 33 might also drive 33 gold value to double to 67. 67/33 transitions to 33/67 with no capital loss, rebalancing back to 67/33 stock/gold again has you holding twice as many shares after prices/values had halved. That will lag across times when stocks do well, decline less across bad stock times. More widely/broadly tend to largely compare to all-stock overall when measured from average to average points (but could selectively be measured peak-trough or trough-peak if one wanted to make a specific distorted sales-pitch angle).
So a investor might opt to align once/year to the iWave indicated cash at that time, or the vwave, or run AIM, or 67/33 to in effect comparable expectancies. One of those will be the worst, another the best, but I suspect that's more a case of luck than specific individual consistent characteristics. Individual AIM wise I haven't measured actual results as part of that analysis, however I feel that falls into the same category. When you buy (or sell) some shares the prior trend might continue, or reverse, again more a case of luck, trading smaller amounts multiple times/year (averaging) rather than once/year (lump) I suspect broadly washes out overall. Constant weighted, such as 67/33 however is inclined to be the worst of the set. At times 50/50 would have been more appropriate, at other times 80/20, adjusting to align with those levels in reflection of current valuations rather than a constant single weighting choice is likely better than not.
Primarily we're looking at risk-reduction benefits. Maybe similar overall rewards to all-stock in the broad sense, but where SWR risk is lowered (worst case SWR is improved). A factor is that data since the 1980's is across a period that started at significant lows. The 1970's saw large scale losses in stocks and a rise to very high inflation and interest rates, a relative (and significant) low. Starting from a low is inclined to yield good/great subsequent outcome. Presently that's somewhat flipped, recent levels might be considered as being at potential highs (more so given recent prior years of 0% interest rates/inflation). AIM and such-like did OK over that low to high era transition and reasonably might be expected to do better during a high to low period.
Fiat currency is a given. Even under gold standard (gold/silver/copper is money) revisions periodically have to be made as things are never consistently stable. Under the gold standard revisions are short/sharp large single points in time adjustments. Under fiat its more a glide-path situation, interest rates, debasement (printing/spending money), taxes etc. are adjusted towards a smoother glide path objective. It's pretty much a given that a dollar today will buy more than that dollar will buy in 20 years time.
More generally I suspect a reasonable choice might be just to apply the vWave once/year portfolio rebalancing as the least effort choice (iWave/MRI being non-public). Or apply AIM if you so prefer. With AIM that opens up applying separate AIM's to different sectors/indexes that is inclined to further smooth out overall portfolio volatility (potentially further lower SWR risk). Failing that, just opt for something like 67/33 yearly rebalanced. Much of investing is more about reducing mistakes. Many individual investors are attracted in at highs, capitulate after losses/lows, mistakes that can be very costly, end up being worse off than if they'd just deposited their money into a interest paying cash deposit account. Outside of that, adding-when-low/reducing-when-high is the counter side to that, and where even if you just achieve market average rewards that average outcome is typically good.
Clive
Market Data 
Markets 
AI
