With $0 commissions, I'm thinking of creating true small-value index. Please talk me out of it

[financeperchance]

Original Fama/French 3 factors here:
https://onlinelibrary.wiley.com/doi/ful ... .tb04398.x

This means buying stocks that are:
1) Smallest decile market capitalization, which (if you eliminate the ones below $50 million) means US stocks with a market cap of only 50m-124m, as found at finviz.com (See table V in that paper)
2) Stocks in the top decile of book-to-market. (Table IV in that paper)
3) No financial stocks, since they were not in the Fama/French set. (See section I of that paper for discussion of this.)

It's impossible to buy an ETF or mutual fund that sticks to these three criteria. Market caps are way too small for this to be practicable for a fund. But an individual certainly can.

Screen for these three things, and like Fama/French did, revisit once a year and sell the things that are no longer in the list and buy the things that have entered it.

Please poke holes in my idea before I take the plunge. Thank you.

[financeperchance]

Wanted to add that looking at the Finviz screener, there are 3,602 US stocks over $50 million in market cap. The total set of companies that are in the intersection of cheapest 10% of P/B plus cheapest 10% of market cap plus non-financials are maybe 100 or so stocks.

[whodidntante]

I'm not going to talk you out of it, but I don't think it's trivial to implement a factor fund well. Please listen to this interview with DFA's head of investing solutions so you can begin to understand the scope involved at least at a pop-science level. Very good insights on DFA's black box by someone who knows what's in it, and why.

Ep 79.
https://tunein.com/podcasts/Business--E ... =136629647
Today on the show we welcome the Head of Investing Solutions at Dimensional Fund Advisors, Marlene Lee. Marlene has a Ph.D. from the University of Chicago where she served as the TA to Eugene F. Farma. She has been at Dimensional for 11 years and is currently also serving as the Co-head of Research, where a big part of her role is communicating what their research team is doing for the advisors and clients who are using their products.

[financeperchance]

I'm not going to talk you out of it, but I don't think it's trivial to implement a factor fund well. Please listen to this interview with DFA's head of investing solutions so you can begin to understand the scope involved at least at a pop-science level. Very good insights on DFA's black box by someone who knows what's in it, and why.

Ep 79.
https://tunein.com/podcasts/Business--E ... =136629647
Today on the show we welcome the Head of Investing Solutions at Dimensional Fund Advisors, Marlene Lee. Marlene has a Ph.D. from the University of Chicago where she served as the TA to Eugene F. Farma. She has been at Dimensional for 11 years and is currently also serving as the Co-head of Research, where a big part of her role is communicating what their research team is doing for the advisors and clients who are using their products.

Thanks! I'll check it out.

[jhfenton]

Ep 79.
https://tunein.com/podcasts/Business--E ... =136629647
Today on the show we welcome the Head of Investing Solutions at Dimensional Fund Advisors, Marlene Lee. Marlene has a Ph.D. from the University of Chicago where she served as the TA to Eugene F. Farma. She has been at Dimensional for 11 years and is currently also serving as the Co-head of Research, where a big part of her role is communicating what their research team is doing for the advisors and clients who are using their products.

Thanks for the link. I wasn't familiar with that podcast, but this latest episode looks interesting. I downloaded it to listen to later.

[ray.james]

Ep 79.
https://tunein.com/podcasts/Business--E ... =136629647
Today on the show we welcome the Head of Investing Solutions at Dimensional Fund Advisors, Marlene Lee. Marlene has a Ph.D. from the University of Chicago where she served as the TA to Eugene F. Farma. She has been at Dimensional for 11 years and is currently also serving as the Co-head of Research, where a big part of her role is communicating what their research team is doing for the advisors and clients who are using their products.

Thanks for sharing it. Very insightful!

[danielc]

Please poke holes in my idea before I take the plunge. Thank you.

Fama and French would tell you that your plan is very risky. They do not see the factors as free money. They see the factors as additional sources of risk. There is a small cap premium, and a value premium, because those companies really are a lot riskier. For example, Larry Swedroe (huge advocate of factor investing) has warned that small cap and value companies tend to precipitously drop in price at the worst possible times. When there is a market correction, suddenly the correlations between these stocks and the market shoot to 1, and these stocks lose more than the rest of the market. So there is a serious danger of a bad outcome at a really bad time. Value companies tend to be in some form of financial distress, and they often do not operate at full capacity (e.g. a company has enough factories to make a million widgets a month but they only make half a million). When there is a bull market, they can quickly make profit because they can easily ramp up production, but in a market downturn there is a real risk that they'll collapse entirely. There are similar risk-based arguments for small cap stocks. As an investor you should be mainly interested in risk-adjusted return. If you are willing to view risk as more than just a standard deviation, and consider the other risks of small caps and value stocks, Fama and French would argue that your risk-adjusted return for small cap and value are not any better than that of the market.

[sometimesinvestor]

Please explain why the hours or days of research and the time needed to make 100+ trades are worth it

[nisiprius]

Not all strategies automatically become good just because of $0 commissions. Is your strategy so fragile that it is good with $0 commission, but would have been bad three years ago at the least-expensive brokers? If not, why weren't you doing it before?

What do you know about liquidity and bid-ask spread for these stocks?

Presumably, these holdings will be part of a larger portfolio that includes other stocks and bonds. Will using a small amount of higher-proof vodka really taste that much better than simply using a somewhat larger amount of the somewhat lower-proof vodka you can buy in convenient mutual fund or ETF form? Have you actually made that comparison?

Once your portfolio is established, how will you add to it? Make twenty, fifty, a hundred transactions every time? How will you sell some of it? Twenty, fifty, a hundred transactions every time?

[stlutz]

Would you be able to do it all in a tax-advantaged account?

If 5 years from now there *is* an ETF that better meets your requirements and wanted to switch to that, or if your requirements change, it gets a lot more problematic from a tax perspective to make the switch in a taxable account if you [hopefully] have a lot of gains.

[PluckyDucky]

Would you be able to do it all in a tax-advantaged account?

If 5 years from now there *is* an ETF that better meets your requirements and wanted to switch to that, or if your requirements change, it gets a lot more problematic from a tax perspective to make the switch in a taxable account if you [hopefully] have a lot of gains.

He could do it on M1Finance in a Roth/Trad IRA. Select the stocks you want once a year for your pie and rebalance it. New money coming in would be auto-allocated per the current pie.

M1Finance's limit per pie is 100 securities.

[lifeisinmirrors]

I think markets are a lot more efficient than they were from 1963-1990 in the F/F data set. Warren Buffet used to find overlooked companies trading at two or three times earnings; nothing is overlooked now when there are more funds than stocks. I think you would be buying a lot of value traps with that strategy.

[Theoretical]

Honestly, I’d just buy RZV and possibly also Franklin’s Microcap Value fund (strict quant non-index) or Aegis Value (extreme micro and cigar-butt value run by a strict Fama-French manager) and call it a day.

I think you could do this but would need some serious Excel-fu to make it a low-maintenance “buy now, sell now” automated system to make sure your deep value stock hasn’t turned into a growth company.

[acegolfer]

Original Fama/French 3 factors here:
https://onlinelibrary.wiley.com/doi/ful ... .tb04398.x

1. That 1992 paper doesn't mention SMB and HML. If you want to see the original FF 3 factor model, here's the 1993 paper: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf

2. These 2 factors require short positions of Big and Low B/M stocks. Are you going to short? Or will you just long Small + High B/M?

[MotoTrojan]

Would you be able to do it all in a tax-advantaged account?

If 5 years from now there *is* an ETF that better meets your requirements and wanted to switch to that, or if your requirements change, it gets a lot more problematic from a tax perspective to make the switch in a taxable account if you [hopefully] have a lot of gains.

He could do it on M1Finance in a Roth/Trad IRA. Select the stocks you want once a year for your pie and rebalance it. New money coming in would be auto-allocated per the current pie.

M1Finance's limit per pie is 100 securities.

I would guess the vast majority of the companies that actually fall in to the OPs requirements would not meet M1's liquidity screening, nor would I want to purchase them blindly (no limit order) myself if I could.

[rkhusky]

The FF portfolios are not meant to be investments. They were developed to be orthogonal bases to explain past stock market returns. The FF decomposition does not make any prediction of future portfolio performance.

[financeperchance]

I think markets are a lot more efficient than they were from 1963-1990 in the F/F data set. Warren Buffet used to find overlooked companies trading at two or three times earnings; nothing is overlooked now when there are more funds than stocks. I think you would be buying a lot of value traps with that strategy.

I think you're right about this. It's funny, the companies when you run the screen are extremely horrible, things like terrestrial radio station companies, whose best days are way behind them and have virtually zero hope of a turnaround. I realized there's no way I'd want my financial future hitched to any of them.

I'd be interested to see what sorts of companies came up decades ago and whether they were that bad. Anecdotally, Buffett and Munger have said several times at the Berkshire meetings that the game used to be a lot easier several decades ago than it is today.

[financeperchance]

Anyway, I just wanted to thank everyone who contributed to this thread. Each of you made good points, and I won't go through with this lamebrained idea.

[nisiprius]

The FF portfolios are not meant to be investments. They were developed to be orthogonal bases to explain past stock market returns. The FF decomposition does not make any prediction of future portfolio performance.

Actually, they aren't orthogonal. I know that because I am fearless about emailing anyone who a) is a professor and b) publishes their email address, and I wrote to Kenneth R. French, asking:

I’m a layperson with trying to understand the basics directly…

I downloaded "Historical Benchmark Returns (Downloadable Files): Fama/French Benchmark Factors, Monthly,” from

http://mba.tuck.dartmouth.edu/pages/fac ... onthly.zip

I imported it into Excel, getting a sheet with 1081 rows, row 2 being headers, row 3 being 192607, down to row 1081 being 201606.

I used the Excel “correl” function, e.g. =CORREL(B2:B1081,C2:C1081).

The correlation between Rm-Rf and SMB came out 0.337, the correlation between Rm-Rf and HML came out 0.2364, and the correlation between SMB and HML came out 0.118.

I expected them all to be zero, because I thought factors were specifically constructed so as to have zero cross-correlations.

What am I misunderstanding?

And Professor French, to his credit, suffered this fool gladly and simply replied:

They are not constructed to have zero correlation.

I didn't try to dig any deeper than that.

[aristotelian]

Would you be able to do it all in a tax-advantaged account?

If 5 years from now there *is* an ETF that better meets your requirements and wanted to switch to that, or if your requirements change, it gets a lot more problematic from a tax perspective to make the switch in a taxable account if you [hopefully] have a lot of gains.

He could do it on M1Finance in a Roth/Trad IRA. Select the stocks you want once a year for your pie and rebalance it. New money coming in would be auto-allocated per the current pie.

M1Finance's limit per pie is 100 securities.

+1, and OP please post your pie to this thread so others can use it. viewtopic.php?t=291548

[Carol88888]

Consider it your "play" money and keep to less than 5% of assets.

It might actually work. I read that Sir John Templeton once took $10,000 and told his broker to buy every stock on the exchange selling for under $1.00. He then went on a trip.

When he came back asked how much he had left and was actually surprised that it did much better than he had hoped.

[rkhusky]

And Professor French, to his credit, suffered this fool gladly and simply replied:

They are not constructed to have zero correlation.

I didn't try to dig any deeper than that.

That's great that he responded. While the factors weren't designed to be orthogonal and are only roughly orthogonal in practice, I would think there would be issues if the correlations were large (>0.7). They are useful for practical application, not for theoretical purity.