Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
These software use geometric averaging of returns. So, if period returns (daily, weekly, monthly) are d1, d2, d3,....,dn, then total-return (TR) is, TR = (d1*d2*d3*......*dn)^(1/n).
For 30 years, with monthly returns, n = 30*12 = 360 only, for daily returns, n = 30*252 = 7,560.
Daily data are more prone to errors due to ex-div date errors and rounding.
So, I am OK with TR, CAGR being close enough.
I would be more concerned about tiny differences among the software that use monthly data (MFO, PV, M*), but there could be minor discrepancies in TR, CAGR with those that use daily data (TestFol. Stock Rover).
BIG differences for SD, Sharpe Ratio, Sortino Ratio, Drawdowns have already been explained.
I mentioned in my post on 8/3/25 that weekly data would be a good compromise (n = 30*52 = 1,560) but that's isn't used by many (StockCharts has Daily, Weekly and Monthly (paid version) views and so are the related data, but that's a charting software, so it wasn't included in my review).
I can predict that all these FREE software will use us, the users, to debug and then flip to subscription software. Morningstar is doing that, PV and StockCharts already charge arm-&-leg for their paid versions.
check out TF's chart for HOSIX vs Morningstar's or stockchart's. according to TF it still hasn't recovered from an April decline and is showing a current ytd gain of about .8% with no new highs where every other place has it at 3.71% with monthly new highs. that's one example but there are others i can't recall right now. i did write to TF but never heard back. then again, maybe i'm reading things wrong.
I see the issue with HOSIX after April. Somehow, no reinvestments after April (before OK). I will also send a bug report via email.
TestFol response was that its data provider isn't great for mutual funds & may change it.
Comments
Fund Performance Metrics (Jan 4, 2016 – Jul 31, 2025)
Fund Name & Ticker CAGR (Compound Annual Growth Rate)
Max Drawdown
Sharpe Ratio
Sortino Ratio
Marshfield Fund (MRFOX) 12.15% -30.88% 0.65 0.94
T. Rowe Price Capital Appreciation (PRWCX) 14.72% -26.54% 0.82 1.21
T. Rowe Price U.S. Equity Research (PRCOX) 13.98% -34.11% 0.73 1.05
Vanguard 500 Index Fund (VFIAX) 14.35% -33.72% 0.76 1.09
https://www.tiingo.com/products/end-of-day-stock-price-data
$30/month
Claims to have data from 1970s, then can have AI write Python code to generate output needed.
TR = (d1*d2*d3*......*dn)^(1/n).
For 30 years, with monthly returns, n = 30*12 = 360 only, for daily returns, n = 30*252 = 7,560.
Daily data are more prone to errors due to ex-div date errors and rounding.
So, I am OK with TR, CAGR being close enough.
I would be more concerned about tiny differences among the software that use monthly data (MFO, PV, M*), but there could be minor discrepancies in TR, CAGR with those that use daily data (TestFol. Stock Rover).
BIG differences for SD, Sharpe Ratio, Sortino Ratio, Drawdowns have already been explained.
I mentioned in my post on 8/3/25 that weekly data would be a good compromise (n = 30*52 = 1,560) but that's isn't used by many (StockCharts has Daily, Weekly and Monthly (paid version) views and so are the related data, but that's a charting software, so it wasn't included in my review).
I can predict that all these FREE software will use us, the users, to debug and then flip to subscription software. Morningstar is doing that, PV and StockCharts already charge arm-&-leg for their paid versions.