TRBCX vs VTSAX/VFIAX - interesting

[keyvaluepair]

I wanted to put this item up for discussion. If you look at 3 month rolling returns for TRBCX vs a fund like VFIAX, it seems that when the market is doing well, TRBCX outperforms VFIAX enough to make up the higher load. When the market does poorly, the opposite happens. This is a bit contrary to the normal religion of not being able to beat the wide market (btw, VTSAX vs VFIAX are mostly identical).

Thoughts? Comments?

[NoStacheOhio]

I wanted to put this item up for discussion. If you look at 3 month rolling returns for TRBCX vs a fund like VFIAX, it seems that when the market is doing well, TRBCX outperforms VFIAX enough to make up the higher load. When the market does poorly, the opposite happens. This is a bit contrary to the normal religion of not being able to beat the wide market (btw, VTSAX vs VFIAX are mostly identical).

Thoughts? Comments?

A. Don't call it a religion. There's plenty of data to back up low-cost index funds as the optimal choice for long-term investing.
B. Nobody said you can't beat the market. Plenty of people beat the market. They're mostly distributed randomly amongst all of the other people trying to beat the market. About what you would expect, really.

[keyvaluepair]

Perhaps the term religion is a bit too strong. Yes, there is plenty of evidence on the value of low cost indexes & I've got quite a bit of vfiax and vtsax myself. However, it seems to me that multiple years of beating the S&P is nothing to be sneezed at.

[NoStacheOhio]

Perhaps the term religion is a bit too strong. Yes, there is plenty of evidence on the value of low cost indexes & I've got quite a bit of vfiax and vtsax myself. However, it seems to me that multiple years of beating the S&P is nothing to be sneezed at.

Religion is based on faith instead of facts. Advocating long-term index investing is based on facts. They're just not comparable things.

Call me when they've beaten the S&P by a statistically significant margin (after taxes & fees) every year for 50 years. I'll even give them two off years.

[simonsez]

Perhaps the term religion is a bit too strong. Yes, there is plenty of evidence on the value of low cost indexes & I've got quite a bit of vfiax and vtsax myself. However, it seems to me that multiple years of beating the S&P is nothing to be sneezed at.

The expense ratio is typically about 15x higher than VTSAX and your window is pretty small.  Still, even when you go back to the summer of 1993 you get a CAGR spread of 1.5% over the S&P500/VTSAX.  So it holds true for that period accounting for the expense ratio, congrats!

This particular period in question has about 3.5 years of bear market and 20 years of bull market.  If you think this will be similar moving forward, you're all set.

Personally I am incredibly lazy so just give me the cheapest thing that captures most of the market with some kind of proportionality baked in (like VTSAX!) that I don't have to pay attention to.  If others invest in something else that has better net returns, that's great!  Something about higher fees just grates against my being. YMMV

[keyvaluepair]

Perhaps the term religion is a bit too strong. Yes, there is plenty of evidence on the value of low cost indexes & I've got quite a bit of vfiax and vtsax myself. However, it seems to me that multiple years of beating the S&P is nothing to be sneezed at.

Religion is based on faith instead of facts. Advocating long-term index investing is based on facts. They're just not comparable things.

Call me when they've beaten the S&P by a statistically significant margin (after taxes & fees) every year for 50 years. I'll even give them two off years.

OK. If we are talking statistics, then lets get precise. Here is the specific analysis that I propose (and will do):

(1)  We have two time series of returns. One is the S&P series and the the other is the one under test viz. PRGSX.
(2) Assuming each observation is drawn from some underlying distribution of returns, we have a random variable F that captures the returns of the S&P 500 and the other, G, is the return of PRGSX after both have been normalized to have a zero cost.
(3) If we test the hypothesis that F is stochastically dominated by G and the hypothesis is rejected at p=0.05 does that do it for you?

kv

[Indexer]

If you look at 3 month rolling returns for TRBCX vs a fund like VFIAX, it seems that when the market is doing well, TRBCX outperforms VFIAX enough to make up the higher load.

TRBCX overweights consumer discretionary. That sector normally shifts with the economy more than other sectors. It does really well when the economy is going well, but it tends to take a bigger hit in recessions.

However, it seems to me that multiple years of beating the S&P is nothing to be sneezed at.

Achoo!

There are thousands of active funds. What are the odds that a coin will come up heads 10 times in a row? Pretty slim, but if you flip it thousands of times I'm sure it will flip heads 10 times in a row at some point. Look hard enough and you will find other active funds that have outperformed the market for a few years in a row here and there. There are active funds that beat their benchmarks, even over time. It happens. No one disputes that. However, it happens rarely enough that it's hard to tell, is it skill or luck?

TRBCX has done well lately, but if you look at a chart of it's whole history it performed pretty similar to the S&P 500 right up until 2009. Even after 2009 it has been hit or miss. There were a few years where it really shined, and several years where it did about the same or worse than the S&P 500.


When it outperforms for multiple decades, then tell me about it. Speaking of such funds...

VPMAX: It has outperformed the S&P 500 and TRBCX over the long term. Over 16 years it has averaged 3% returns above the S&P 500. The higher cost original version, VPMCX, has outperformed the S&P 500 by about 2.5% per year going back to 1984.

This is one of those few cases I think it might actually be skill, but VPMAX breaks all the rules that hinder other active funds. It has very low turnover, lower than it's comparable index fund, VIGAX. The managers make decisions with the long term in mind, and don't bother themselves with the traditional short term quarterly mindset. It has very low costs for an active fund, less than half of the cost of TRBCX. It's also closed... When active funds do really well they attract more money. Most managers are greedy so they keep taking new funds, and eventually they run out of great ideas for investing the new funds so they start investing in stocks they would have passed over before(or.. their luck runs out). That normally leads to lower future performance.  Primecap closed to new investors to prevent this from happening.

Finding a fund like Primecap before it closes is just about impossible... which brings us back to index funds. There is also no guarantee Primecap will continue to outperform. 30+ years would be a whole lot of luck, but statistically it is possible.

[keyvaluepair]

Do you agree with the proposed analysis or not. Let's be quantitative. I'll get my analysis done this weekend. What is your analysis proposal precisely?

[keyvaluepair]

OK. I did the analysis. Here are the steps:
(1) Computed the rolling returns starting 2002 to current for both VTSAX and TRBCX. These were then adjusted for the different fund costs.
(2) I wanted to compare the return to see if TRBCX (adjusted) had similar returns to VTSAX (adjusted).

Here is what I found:

Over the 2002-2017 period, there was no significant difference between the two.

For example, comparing the probability density functions (a smoothed histogram) of the two, they look almost identical.


The Kruskal-Wallis test for median comparison yields no significant difference:

 kruskal.test(list(vtsax.adjusted.returns, trbcx.adjusted.returns))


Kruskal-Wallis rank sum test

data:  list(vtsax.adjusted.returns, trbcx.adjusted.returns)
Kruskal-Wallis chi-squared = 0.3638, df = 1, p-value = 0.5464


So, I thought about looking at the relative performance for the last 8 years.

Here is the comparison of the probability density functions:


More promising? No:
 kruskal.test(list(vtsax.last8.years, trbcx.last8.years))

Kruskal-Wallis rank sum test

data:  list(vtsax.last8.years, trbcx.last8.years)
Kruskal-Wallis chi-squared = 0.67087, df = 1, p-value = 0.4127

Consequently, my initial hypothesis of there being a difference is wrong! But without actually testing it, we can't say. I'll do the same for vpmax out of curiosity.

[AZryan]

Ignoring the swamp of your extremely extensive technical research, because past performance, cost and description of TRBCX is (as explained by others here already) simply unconvincing and unimpressive...

Here's a far better VTSAX substitute... VIMAX. Vanguard's Midcap Fund.

Long-term, it trounces the Total Market (and TRBCX). It's still indexing, and it's fee is nearly as dead low as the Total Market. It's essentially a LargeCap and Midcap fund, whereas the Total Market is really a MEGAcap and Largecap fund. This means it's still a fund dominated by major companies we all know (just as VTSAX/VFIAX are), and not flash-in-the-pan types in a Smallcap Fund. It's a Goldilocks.

Slightly more volatile than the Total Market, but essentially follows it perfectly (except when it mostly ignored the dotcom crash). It's lagged the Total market over the past ~3years, but not by much (making it a good time for new people to shift in -in a rebalancing-type way, not a market-timing way), and long-term more than made up for it.

It's my primary fund. Total Market is my smaller, secondary fund. And the rare times the Total Market beats it, it lets me rebalance a little without needing any Bonds. The combo of those two halves the impact/risk of companies like Apple and Netflix, etc. from holding SO MUCH sway over the entire market (look up just how VERY different that threat is now compared to the top companies of the past).

You'll basically end up with one 'set it and forget it' VERY broad Largecap Index that you might rebalance a tiny bit every several years (if at all).

And for all the active funds out there... T Rowe Price's Media/Telecom and Global Tech funds (PRMTX/PRGTX) have far more than paid for their near 1% fees over many years. I personally believe a focus on technology will always be 'where it's at' going forward. Long, consistent records through good times and bad. 5% of each, mixed with 90% Total Market, and you ought to beat the market regularly with very little added risk.

A combo of Mid/Total/PRMTX/PRGTX has proven to do extremely well.

[keyvaluepair]

Ignoring the swamp of your extremely extensive technical research, because past performance, cost and description of TRBCX is (as explained by others here already) simply unconvincing and unimpressive...

Here's a far better VTSAX substitute... VIMAX. Vanguard's Midcap Fund.

Long-term, it trounces the Total Market (and TRBCX). It's still indexing, and it's fee is nearly as dead low as the Total Market. It's essentially a LargeCap and Midcap fund, whereas the Total Market is really a MEGAcap and Largecap fund. This means it's still a fund dominated by major companies we all know (just as VTSAX/VFIAX are), and not flash-in-the-pan types in a Smallcap Fund. It's a Goldilocks.

Slightly more volatile than the Total Market, but essentially follows it perfectly (except when it mostly ignored the dotcom crash). It's lagged the Total market over the past ~3years, but not by much (making it a good time for new people to shift in -in a rebalancing-type way, not a market-timing way), and long-term more than made up for it.

It's my primary fund. Total Market is my smaller, secondary fund. And the rare times the Total Market beats it, it lets me rebalance a little without needing any Bonds. The combo of those two halves the impact/risk of companies like Apple and Netflix, etc. from holding SO MUCH sway over the entire market (look up just how VERY different that threat is now compared to the top companies of the past).

You'll basically end up with one 'set it and forget it' VERY broad Largecap Index that you might rebalance a tiny bit every several years (if at all).

And for all the active funds out there... T Rowe Price's Media/Telecom and Global Tech funds (PRMTX/PRGTX) have far more than paid for their near 1% fees over many years. I personally believe a focus on technology will always be 'where it's at' going forward. Long, consistent records through good times and bad. 5% of each, mixed with 90% Total Market, and you ought to beat the market regularly with very little added risk.

A combo of Mid/Total/PRMTX/PRGTX has proven to do extremely well.

I think that you may missed my main point. If you don't do the analysis, then what you have left is opinion. The analysis above is not very technical statistically. It merely checks if the two random variables - one for VTSAX and the other for TRBCX have statistically significant differences - in fact if you look at the density of TRBCX for the last 8 years, it skews right relative to VTSAX. But not enough to pass a non-parametric test.

I had an opinion myself, but that turned out to be unfounded on the basis of analysis - I don't have to take anyone's opinions on it, since the other posts were not analytically backed. Not being arrogant, just factual.

[AZryan]

You're not going to get very far with posts like that -that's, of course, my opinion.

The fact is, your timeframes are narrow and your extreme use of obtuse jargon is not likely to be helpful to you or anyone trying to help you. You might be able to pull it off if you could prove it's made you a financial wizard and you've been tremendously helpful to others here. As it is, it looks pretty silly and not 'just factual'.

You don't seem to care about the facts suggesting VIMAX is a much better VTSAX substitute. That's disappointing, but maybe someone else will be helped by that info and bother looking into it.

[NoStacheOhio]

If you don't do the analysis, then what you have left is opinion.

I think you're overlooking the prior analyses/consensus available. Just because I didn't do the math myself doesn't make it opinion.

If, for example, the conclusions we've come to as a result of the Trinity Study eventually get knocked down, it won't be because one person on the internet did analysis. It'll happen in a pretty transparent way, with lots of experts publishing their findings, and eventually generating a new consensus.

[keyvaluepair]

Again, please don't mistake what I say for arrogance. What I am saying is that I have not seen an analysis that suggests that VIMAX is indeed better than VTSAX. In fact the point is that indexing works is supported in the case above. I'd be glad to do the analysis for VIMAX when I can.  Getting the returns data is most of the work, unfortunately.

@azryan: I don't know what you mean by "jargon" since I was simply trying to make the details of the analysis clear. It is very easy to lie with statistics. For example, if I had simply stopped with comparing the median returns of the two funds - even after adjustment TRBCX is higher. The question is whether it is higher, without doing a statistical hypothesis test.

That is what was done, which means that others can do it themselves, which is the essence of peer review.

[Radagast]

We don't think you are arrogant. We think you are ignorant. You obviously have no idea what you are doing.

Read:
Bernstein The Four Pillars of Investing
Berkin & Swedroe Your Complete Guide to Factor-Based Investing
Bogle Common Sense on Mutual Funds

Then play around with https://www.portfoliovisualizer.com/factor-analysis?s=y&regressionType=1&symbols=TRBCX+VFIAX+VIMAX&startDate=1%2F1%2F2002&endDate=11%2F30%2F2017&factorDataSet=0&marketArea=0&factorModel=4&useHMLDevFactor=false&includeQualityFactor=false&includeLowBetaFactor=false&fixedIncomeFactorModel=0&__checkbox_ffmkt=true&__checkbox_ffsmb=true&__checkbox_ffsmb5=true&__checkbox_ffhml=true&__checkbox_ffmom=true&__checkbox_ffrmw=true&__checkbox_ffcma=true&__checkbox_ffstrev=true&__checkbox_ffltrev=true&__checkbox_aqrmkt=true&__checkbox_aqrsmb=true&__checkbox_aqrhml=true&__checkbox_aqrhmldev=true&__checkbox_aqrmom=true&__checkbox_aqrqmj=true&__checkbox_aqrbab=true&__checkbox_trm=true&__checkbox_cdt=true&timePeriod=2&rollPeriod=36&marketAssetType=1&robustRegression=false
to see if it satisfies your need for statistical analysis of what fund did better, why, and how significant it was based on rigorous academic research. Then, if you still think what you are doing is useful, carry on.

[keyvaluepair]

We don't think you are arrogant. We think you are ignorant. You obviously have no idea what you are doing.

Read:
Bernstein The Four Pillars of Investing
Berkin & Swedroe Your Complete Guide to Factor-Based Investing
Bogle Common Sense on Mutual Funds

Then play around with https://www.portfoliovisualizer.com/factor-analysis?s=y&regressionType=1&symbols=TRBCX+VFIAX+VIMAX&startDate=1%2F1%2F2002&endDate=11%2F30%2F2017&factorDataSet=0&marketArea=0&factorModel=4&useHMLDevFactor=false&includeQualityFactor=false&includeLowBetaFactor=false&fixedIncomeFactorModel=0&__checkbox_ffmkt=true&__checkbox_ffsmb=true&__checkbox_ffsmb5=true&__checkbox_ffhml=true&__checkbox_ffmom=true&__checkbox_ffrmw=true&__checkbox_ffcma=true&__checkbox_ffstrev=true&__checkbox_ffltrev=true&__checkbox_aqrmkt=true&__checkbox_aqrsmb=true&__checkbox_aqrhml=true&__checkbox_aqrhmldev=true&__checkbox_aqrmom=true&__checkbox_aqrqmj=true&__checkbox_aqrbab=true&__checkbox_trm=true&__checkbox_cdt=true&timePeriod=2&rollPeriod=36&marketAssetType=1&robustRegression=false
to see if it satisfies your need for statistical analysis of what fund did better, why, and how significant it was based on rigorous academic research. Then, if you still think what you are doing is useful, carry on.

I did read the Swedroe book and the Fama papers. Abuse is always a sign of lack of an argument.

[Radagast]

If you actually read Swedroe and Fama and you are stilling doing statistical analyses of the past price or total returns of an actively managed mutual fund to see if it will be a good investment for the future, then by all means carry on.

That wasn't abuse. That was truth.

[keyvaluepair]

There was no evidence of either autocorrelation, trend or heteroscedasticity in either series. Explain why a stationary analysis is not sufficient.

[Indexer]

Do you agree with the proposed analysis or not. Let's be quantitative. I'll get my analysis done this weekend. What is your analysis proposal precisely?

No, I don't agree with your analysis. Sure, let's be quantitative.


Summary of everything said so far by most of us:  Your sample size of data is too small to come to ANY conclusions. With any investment it is possible to just get lucky... or unlucky... in favor... or out of favor. Researching how one active fund outperformed over a short period of time means absolutely squat. Having a few really good years the past decade doesn't mean anything. With that short of a time frame, 1 or 2 really stellar years because their largest holding is... Amazon!... will skew the whole data set heavily towards the positive. When you are measuring active managers how do you separate 'luck' from 'skill'?  Time...  You just need more and more data. You don't have enough data so no one here is going to be persuaded by your calculations.

This is why I gave VPMAX/VPMCX as an example. It has outperformed for over 30 years. That is a 'decent' sized data set. It is possible for an active fund to outperform over long time frames, but it's very rare.

Read:
Bernstein The Four Pillars of Investing
Berkin & Swedroe Your Complete Guide to Factor-Based Investing
Bogle Common Sense on Mutual Funds

This, and I'll add a Random Walk Down Wall Street and Intelligent Investor. Personally, I wish the first investing book I ever read was Random Walk. The name is fitting. It kind of teaches you the basics, enough to be dangerous, then it explains everyone who invests knows that so really you know nothing, and then eventually ends with why you should just used index funds.