Bogleheads.org

Hyperchicken wrote: ↑Wed Mar 29, 2023 12:26 pm It's hard to comment on generalities. Did they substantiate that claim in any way?

Basically said that with MF you pay capital gains every year as a function of other people’s sales of shares but with ETF only pay capital gains when your own shares are sold. I find it hard to believe there are any free lunches, so I’m curious. That being said, I know there is a trend for MFs to convert to ETFs, and I think there has to be something positive behind that trend.

Dave

The main reason I ask the question is I have someone telling me that ETFs are far superior with regard to taxes. Any comments from a tax perspective?

Dave

Whether we’re talking total stock market investors (VTI v. VTSMX) or factor junkies (AVDV v. DISVX), is the ETF structure just plain superior to the MF structure for buy and hold investors? I’m assuming similar holdings characteristics when comparing an ETF and a MF. Are there cases where the MF is preferred? Thanks,

Dave

Short answer is no. Sequence of Risk returns are tied to volatility. Value stocks have equal or higher volatility than the total market. My short answer is maybe but I wouldn't get my hopes up too high. Value stocks would have to be in the context of a diversification across factors strategy that Larry Swedroe has talked about. Diversification across factors worked well during the 2000-2002 bear market, failed during the 2008-2009 bear market and financial crisis, and helped during 2022. No good diversification strategy works well all of the time. Important to remember that size and value are considered risk factors with an associated expected premium. Thus for comparison to a TSM portfolio, one needs to guesstimate a tilted portfolio with...

I think the key to minimizing SORR in this case is applying the value tilt together with decreasing the overall equity allocation a bit. This is a move in the direction of risk parity, more evenly spreading risks across market factor, value, term.

Dave

I think the typical TSMer views diversification as stocks/bonds and US/Int. This book views diversification as across unique and independent sources of risk and expected return. Has it compelled some of us to look at diversification differently in the low expected return environment?

Dave

What if John Bogle had been born in Japan?

Dave

Apathizer wrote: ↑Thu Mar 16, 2023 11:23 am

Random Walker wrote: ↑Thu Mar 16, 2023 11:12 am

rockstar wrote: ↑Wed Mar 15, 2023 4:40 pm Is international really a thing post 90s globalization? Can’t I buy Apple and say I’m globally diversified?

Globalization also I think had a surprisingly large effect in the 1700’s.

Dave

It seems many didn't seem to actually watch the video since he addressed it specifically. No, investing in large companies with global exposure isn't the same as investing in different markets.

I haven’t seen the vide yet, but I certainly will shortly. I’m a pretty Big Ben Felix fan.

Dave

While Chinese companies may post good returns, it doesn't appear the largess is widely dispersed among shareholders. I suppose 8.4% of a small number is an even smaller number. https://www.morningstar.ca/ca/news/225163/china-stocks-the-road-to-nowhere.aspx https://www.morningstar.com/articles/1085524/chinese-stocks-what-went-wrong All markets go through periods of poor performance. The US has had very poor sequences throughout its history that has resulted in returns being worse than Bonds or other equity markets. China is no exception. That does not mean that the expected return on Chinese stocks today is low, or that returns throughout any predefined period are guaranteed to be low. And a 2-3% allocation for a standard globally diversifi...

While Chinese companies may post good returns, it doesn't appear the largess is widely dispersed among shareholders. I suppose 8.4% of a small number is an even smaller number. https://www.morningstar.ca/ca/news/225163/china-stocks-the-road-to-nowhere.aspx https://www.morningstar.com/articles/1085524/chinese-stocks-what-went-wrong All markets go through periods of poor performance. The US has had very poor sequences throughout its history that has resulted in returns being worse than Bonds or other equity markets. China is no exception. That does not mean that the expected return on Chinese stocks today is low, or that returns throughout any predefined period are guaranteed to be low. And a 2-3% allocation for a standard globally diversifi...

rockstar wrote: ↑Wed Mar 15, 2023 4:40 pm Is international really a thing post 90s globalization? Can’t I buy Apple and say I’m globally diversified?

Globalization also I think had a surprisingly large effect in the 1700’s.

Dave

Regarding International (sub) performance, perhaps someone can help me figure this math problem out: From April 30, 2007 (the date I started investing in a taxable account) to February 17, 2023 (the date I last checked), Vanguard Total Stock Market (VTSAX) gained 275.90% while Vanguard FTSE Ex-US (VFWAX) gained 51.44%. (Figures are from Morningstar data). And if I am reading the Morningstar chart correctly, $10,000, invested on April 30, 2007, would now be worth worth $37,590.26 if invested in VTSAX, and $15,099.25 if invested in VFWAX. That’s quite a difference in returns. But here is my first question: Assuming that VTSAX stays stagnant (ie, returns 0% going forward and the value of one’s holding in it stays at $37,590.26), how much woul...

I diversify internationally and have for a long time, but geopolitical risk seems on the rise and I question if some of the countries in which I invest are committed to economic development. And I wonder if that is the thing explaining the luck that Felix cites. Perhaps my losses in Russian energy stocks is short cited, but China too may purse such a road. They are pretty explicitly signaling they will and then what. So in a bifurcation of the global economy, will international diversification be desirable. There is that risk and seemingly little reward for taking it. I mean what might be in the interest of an autocrat may be very different from what's in the interest of the companies in that country. So Felix did nothing to answer the que...

Was listening to Kristy and Bryce from the popular FIRE podcast Millenial Revolution and they made a few statements about their strategy that i'm having trouble understanding. Here is the clip: https://www.youtube.com/watch?v=sxRt2fAzRi0&t=736s First they claim that if you live off the dividends then you can eliminate the sequence risk. How is this possible? Then they claim that if you just spend the dividends then you're not selling anything - however my understanding of dividends is that it comes off the NAV so getting dividends and spending it is functionally equivalent to NOT getting dividends and then selling shares in that amount. Their suggestion is to invest more in International to increase dividend yield because the US market...

But I do think they are rules based and have no individual security selection or market timing. Help me understand something please. Yes, quant funds have algorithms and rules-based processes that allow them to trade to a consistent strategy, but those human-built algorithms still produce buys and sells of individual securities, don’t they? I don’t see how the “human or algorithm” question relates to the “passive or active” question. It seems like the quant guys still are picking individual stocks as opposed to buying the whole basket. Not true? My understanding is that they are buying stocks according to specific characteristics, but if two stocks are identical by their metrics, they are indifferent to which one they buy. Dave

It's possible to see a legitimate grey area around the word "passive." Traditionally, people describing DFA's "classic" funds like the DFA US Micro Cap Portfolio and the DFA US Small Cap Value Portfolio--I think DFA itself although I haven't been able to track it down--describe them as "passive funds, though not index funds." DFA is often thought of as a champion of "passive" investing, and a rival to Vanguard. And I am saying that according to the prospectus for this fund, it is Just Plain Active. There's no suggestion of it being rules-based, and the prospectus specifically says in so many words that it is actively managed. A Just Plain Active fund from DFA seems like a change in direction, althoug...

Another way to think about it: passive seeks beta (the cap weighted return of the market). Any investment strategy that seeks alpha is an active strategy. You and I are basically on the same page I think. I would say though that there is a beta for the market factor, beta for size, beta for value, etc. Past sources of alpha, once discovered, get turned into a beta. And it’s because these funds effectively access these other betas in an agnostic formulaic fashion that I think one can consider them passive. That’s why I say the funds are sort of their own indexes: you get what the factor loadings give you . It comes down to following construction rules. Larry Swedroe’s book The Incredible Shrinking Alpha details the phenomenon of converting ...

I hope we can agree that this DFA product is not "passive?" I agree (even tho there is no agreed definition of “passive”). What do you mean by that? Index funds follow a “investable index” managed by a independent 3rd party. A index is a portfolio constructed by predefined rules. There is nothing magical or special about it. There are SEC marketing rules around this. All index funds are passive, but not all passive funds are index funds. Dave Can you give me a example? Yes. The DFA equity funds I would consider passive. Fama’s definition of passive is no market timing and no individual security selection, and these funds are passive in that sense. They are formulaic, rules based, agnostic to what individual securities are bought....

alex_686 wrote: ↑Wed Mar 01, 2023 9:14 am

Gaston wrote: ↑Tue Feb 28, 2023 8:04 pm

nisiprius wrote: ↑Tue Feb 28, 2023 4:09 pm I hope we can agree that this DFA product is not "passive?"

I agree (even tho there is no agreed definition of “passive”).

What do you mean by that? Index funds follow a “investable index” managed by a independent 3rd party. A index is a portfolio constructed by predefined rules.

There is nothing magical or special about it. There are SEC marketing rules around this.

All index funds are passive, but not all passive funds are index funds.

Dave

Recently read an Asness essay where he adamantly states that strategies such as this are active. Nonetheless I like viewing these formulaic methods as passive. The strategies have no individual stock picking or market timing, are rules based, and replicable.

Dave

Anon9001 wrote: ↑Tue Feb 28, 2023 11:55 am I have skimmed through this thread but MCQ I noticed you are assuming that Geometric Mean= Arithmetic Mean-Variance/2 for these calculations.
Now I don't know if you know this formula is just an approximation and its only exact if the Returns Distribution is Normally Distributed. Its close enough for most cases but it tends to get quite far away from the actual Geometric Mean if the returns distribution is deviating heavily from Normal Distribution.

So is the variance drain larger as distributions deviate more from normal? Does the equation perhaps underestimate the benefits of diversification for investors? Thanks,

Dave

I think a reasonable summary of a lot in the two curves above is as follows. In equity heavy portfolios, the portfolio risk is so dominated by the equity allocation that may as well take on the duration risk and associated premium of longer bonds. The increased risk on the bond side does not appreciably affect the overall portfolio risk but can add a bit to return. In much less aggressive portfolios with substantially lower equity allocations, it is more efficient to limit term risk to intermediate. I believe Larry Swedroe may have written on this.

Dave

delamer wrote: ↑Mon Feb 20, 2023 11:41 am Probably yes.

But is achieving that less volatile portfolio (with the same expected return) possible?

High tilt / lower overall equity allocation?

Dave

abc132 wrote: ↑Mon Feb 20, 2023 11:28 am 1. Correlations change over time
2. SWR's are based on low percentile outcomes

Combining 1) and 2) you can't plan for much if any benefit unless you are utilizing guaranteed payments.

Once again the guaranteed payment is the strongest diversifier. The 40% bonds in a 60/40 portfolio are very powerful.

Potential for annuities to increase safe spending levels in retirement?

Dave

I'm not as experience in such things as you may be; but, as for me the answer would be probably not at this stage of the game. I've been retired for more than 10 years and have been saving / investing since I was 25. (late start). Last year I lost ~20% of the value of my portfolio. There's chatter that I the market could lose another 20% this year. (I know no one knows nothing, but still ...). How high of a withdrawal rate do you think modestly higher should be? I’m thinking modest increases like from 3% to 3.5% or 4%: small differences in percentage terms, but could be significant in absolute freedom to spend terms. Dave That is a pretty massive increase by my standards, not a modest increase. I thought the discussion would be about somet...

RadAudit wrote: ↑Mon Feb 20, 2023 9:17 am I'm not as experience in such things as you may be; but, as for me the answer would be probably not at this stage of the game. I've been retired for more than 10 years and have been saving / investing since I was 25. (late start). Last year I lost ~20% of the value of my portfolio. There's chatter that I the market could lose another 20% this year. (I know no one knows nothing, but still ...).

How high of a withdrawal rate do you think modestly higher should be?

I’m thinking modest increases like from 3% to 3.5% or 4%: small differences in percentage terms, but could be significant in absolute freedom to spend terms.

Dave

We have have talked about portfolio efficiency, variance drain, sequence of returns risk, difference between Ari Mean and Geo Mean. Do people believe that a less volatile portfolio with same expected return could support a modestly higher withdrawl rate? Personally, I think so.

Dave

Where a quant could help is to crunch the numbers and crank the formulas to get an idea of the right amount of diversification. In other words, how many asset classes do you need? If you factor invest, what are the best factors to load on, and how much load for each factor? Being a cautious investors, I don't have extreme tilts but what I am doing with my factor bets might not be enough to make a difference. What I hope Professor McQ will do is to give us an idea of what a more optimal portfolio would look like. If you are using volatile non-correlating asset classes with similar returns to boost return and reduce volatility, what is the optimal number to use? Two? Three? Four? I suspect the correct answer is three. During my annual Nedsai...

2. Historical correlations of factors are significantly less than his 0.5 assumption. For example some rough numbers are Size to Market 0.3 Value to Market -0.3 Size to Value 0.0 Momentum to Market -0.2 Momentum to Value -0.2 >>>This is crucial. Returns from M-diversification are going to be much greater if correlation among equivalent risk assets will range from -0.20 to +0.30. But my copy of the SBBI shows small stocks and large stocks correlated at 0.79 >>>Do you have a source or link for these correlations? Are they annual, or ...? I got the correlations from Chapter 9 Larry Swedroe’s Factor Book. (I think just a few charts in Chapter 9 alone are worth the price of the book) Realize though that these are the correlations of pure long-s...

Northern Flicker wrote: ↑Thu Feb 16, 2023 1:30 am

Random Walker wrote: Yes I mostly disagree with him. I do agree with him that the Ari Mean is what it is and the Geo Mean is what it is; and that we only eat Geo Means. That being said, to intelligently determine our asset allocations, we need to think about expected returns, expected volatilities, and how different potential portfolio components mix in a portfolio. Looking forward, our expected return of an asset is an Ari Mean.

Expected volatility is not a property of a probability distribution. It is imperative to define such a concept before using it. And expected return is not the annual arithmetic mean of a long-term return.

So if you were to generate inputs for a Monte Carlo simulation what would you use?

Dave

McQ clearly knows more than me, but I’ll make a few points. 1. More equal weighting, lower correlation, greater volatility improves diversification benefit. 2. Historical correlations of factors are significantly less than his 0.5 assumption. For example some rough numbers are Size to Market 0.3 Value to Market -0.3 Size to Value 0.0 Momentum to Market -0.2 Momentum to Value -0.2 3. Bonds generally uncorrelated with market but there is a tendency for the correlation to turn strong negative when really need it to (not last year) 4. Can keep expected return of a TSM/TBM portfolio constant by tilting to higher expected return equity factors, decreasing overall equity allocation, increasing exposure to safe bonds. This change would most likely ...

This is an exercise that I have done and think is worthwhile for people pondering these things. In Excel, set up two random return series with the same average return, but different standard deviations. Run the series for 20-30 iterations representing a 20-30 year investing career. See how often the less volatile return series accumulates to a greater amount. It will I think in every instance. And it’s because Geo Mean = Ari Mean - 1/2 SD^2. From what I remember, decreasing SD by around 2%, from say 16 to 14 or 14 to 12, results in Improvements to Geo Mean about the same size as many of the expense ratio differences we discuss on this site, maybe 0.2-0.4%. Of course the huge trick is finding diversifying asset classes that are low correlat...

This is an exercise that I have done and think is worthwhile for people pondering these things. In Excel, set up two random return series with the same average return, but different standard deviations. Run the series for 20-30 iterations representing a 20-30 year investing career. See how often the less volatile return series accumulates to a greater amount. It will I think in every instance. And it’s because Geo Mean = Ari Mean - 1/2 SD^2. From what I remember, decreasing SD by around 2%, from say 16 to 14 or 14 to 12, results in Improvements to Geo Mean about the same size as many of the expense ratio differences we discuss on this site, maybe 0.2-0.4%. Of course the huge trick is finding diversifying asset classes that are low correlat...

This is an exercise that I have done and think is worthwhile for people pondering these things. In Excel, set up two random return series with the same average return, but different standard deviations. Run the series for 20-30 iterations representing a 20-30 year investing career. See how often the less volatile return series accumulates to a greater amount. It will I think in every instance. And it’s because Geo Mean = Ari Mean - 1/2 SD^2. From what I remember, decreasing SD by around 2%, from say 16 to 14 or 14 to 12, results in Improvements to Geo Mean about the same size as many of the expense ratio differences we discuss on this site, maybe 0.2-0.4%. Of course the huge trick is finding diversifying asset classes that are low correlate...

This subject, variance drain, is something I think about a lot. Geo Mean = Ari Mean -0.5*(SD^2). I strongly believe in diversifying across unique and independent sources of risk and expected return. I split equities evenly US and Int, heavily tilt to size and value, decrease overall equity exposure and increase exposure to safe bonds, and have added alternatives. I believe lack of portfolio efficiency is a real cost, and the incremental improvements to portfolio efficiency come with increasing marginal costs. Each individual needs to decide where to draw the line between the potential benefits of improved portfolio efficiency and the certain costs associated with reaching for that potential. Dave Dave, I am just not enough of a mathematici...

McQ wrote: ↑Tue Feb 14, 2023 10:04 pm
Postscript

Very, very few ordinary investors set up Markowitz’ formula in a spreadsheet, and run sensitivity analyses against it, as I’ve done for this thread (after 30 years of teaching MBAs it is hard for me to proceed in any other way).

But I think we can qualitatively take advantage of these results very nicely. The typical Boglehead 60/40 TSM/TBM portfolio is 85-90% dominated by a single factor, market beta. A move away from this dominance is a move towards improved portfolio efficiency.

Dave

If you're not different significantly from the average investor (weighted by portfolio size or trading volume, not the average person who invests) in an articulable way, the market has already done the work for you and produced a portfolio that can be appropriate for you. It's computed the various benefits from diversification and other elements of portfolio construction. Fans of size and value usually have reasons that would apply to pretty much everyone. If you are sufficiently different, then adjustments can be appropriate. How to gauge differences from the average investor (as defined) is not trivial, but if your reasons for picking independent risks are applicable to a wide swathe, then differences are not likely to be massive. This i...

In a world of efficient markets and rational investors, not only would correlations increase, but the "better portfolio" would become the cap-weighted market portfolio, where the better portfolio was some non-cap weighted mixture (VTI plus international or whatever). The process would be buying the diversifier, driving up its price and selling everything else, decreasing those prices, all the while increasing correlations. The process would continue until equilibrium is reached. Market efficiency is most probably a necessary condition to this process, but rational investors is much less clear. You certainly don't need everyone to be rational. All this suggests that the search for low correlation diversifiers that will improve a p...

This subject, variance drain, is something I think about a lot. Geo Mean = Ari Mean -0.5*(SD^2). I strongly believe in diversifying across unique and independent sources of risk and expected return. I split equities evenly US and Int, heavily tilt to size and value, decrease overall equity exposure and increase exposure to safe bonds, and have added alternatives.
I believe lack of portfolio efficiency is a real cost, and the incremental improvements to portfolio efficiency come with increasing marginal costs. Each individual needs to decide where to draw the line between the potential benefits of improved portfolio efficiency and the certain costs associated with reaching for that potential.

Dave

Nathan Drake wrote: ↑Sat Feb 04, 2023 9:16 pm This is sequence of return risk in action. Withdrawing is much more severe than contributing through a drawdown.

It's why diversification is so critical in retirement.

And what I find is that many people seem to not appreciate that diversification may well be most important over the short run; and the importance is amplified in that period around early retirement. For example, if someone retired within the last year or two, he would have blunted his withdrawal pain with a value tilt.

Dave

MnD wrote: ↑Fri Feb 03, 2023 8:07 pm I look at swings in other numbers as well.

Dropped 50 pounds since retirement. Blood work numbers are equally if not more improved.
Running 25 miles per week retired plus serious cross-training - versus 0 miles running and mediocre cross-training prior to retirement.
Averaging 75 leisure travel days per year and 2023 is going to top 100 days. Versus maybe 20 days when working.
We have more wealth now than when we retired in late 2018 which bracketed two or three bears markets and a global pandemic.
12 of my friends, close co-workers or extended family in their 50's or 60's dead since i retired in late 2018. None had any retirement.

What age are you doing that at? I’m 60 and thought running outdoors was over for me.

Dave

Completely agree. Which is why at retirement I will have a TIPS LMP that meets essential expenses that SS/pension will not. If that means my overall assets are at 25/75, so be it. The last year made me reevaluate my future risk tolerance. The thing with equity risk is that when a big decline happens there is no guarantee on the timeframe in which it bounces back. Imagine being reliant on a 60:40 portfolio in 2008 to cover essentials and the market doesn't start to bounce back in 2009 and you spend years pulling from a portfolio that has declined 30%. I find that prospect unsettling even if it has been historically unlikely even though I was 100% stocks (being15 years younger helps!) and had no issues buying into the decline. I completely a...

daleddm wrote: ↑Fri Feb 03, 2023 2:27 pm
Amen. Keeps me going to work despite PRObably (almost certainly) being able to quit. Ye olde "one more year" syndrome. And yeah, I read every thread on why and how folks finally come to conclude it's OK to retire.

One more year can have a big positive effect I think

Dave

So, during the last crash, at the low point, we were down about $250,000 from our all-time portfolio high point in Dec 2021. So, I updated my spreadsheet last night since it was the beginning of the month, and I was quite pleased to see that we're only $50,000 down today. BUT... I'm still working, and that includes all the money we've saved over the past year... So it's not accurate as far as portfolio returns. It includes the money I've invested in my 401k and the match from my employer over the year. Plus it's just nominal numbers... Not accounting for inflation at all. YET... It still makes me feel pretty good to see we are almost back to our all-time portfolio high. But, I was thinking that if I was retired this last year, and was PULL...

BenS wrote: ↑Fri Jan 27, 2023 7:58 pm
What are your thoughts?

Adds conviction to my heavily tilted plan. Persistent over time, pervasive across markets. Data shows presence after publication of the factors. Out of sample data both temporally and geographically supports belief.

Dave

I don’t benchmark my portfolio. I’m apparently a huge factor junkie. But I do believe markets are highly efficient and thus my passive funds do no market timing or individual security selection. I strongly believe that I get what my factor exposures give. The factors I’m exposed to are overwhelmingly market, size, value, quality, momentum, term. How I would do relative to any benchmark depends on how my collection of factor exposures compares to some benchmark’s factor exposures.
Now that being said, I can’t help but look every day at how Int, size, value, bonds compared to VTSMX. Just sort of a game I play with myself. Admittedly way more often than not my highly torqued funds have lost that game 2009-2021ish

Dave

To me it was always intuitive that small caps should do better, cause they can more quickly double etc their profits. A lemonaid stand can quickly double its profits, which something like Walmart can't do. But I think the argument that counters mine is efficient market theory. The market knows the stand can double its profit so the market prices that into the price of the lemonaid stand stock. Even if profits double, the price of the stock does its own thing in anticipation and as time goes on. So is my only way of being right that small cap is better, due to mistakes or inefficiencies in the market? If a doubling of profit isn't factored into the stock price, there definitely could be a small premium. But is there more to it beyond ineffi...

Yes. Value stocks and especially small value have a wider dispersion of potential outcomes than the market. For instance over rolling decade periods expected annual return might be something like 4-10% for the market and 2-15% for small value. The longer the time period the more likely small value is to out-perform the market, but in the short-term there's more risk. However, SV is imperfectly correlated with the market, so I would argue provides beneficial diversification. On it's own SV is riskier though. This is a good example of why diversification across unique and independent sources of return is likely even more important the shorter one’s time period is. I believe that when any source of risk/return can perform poorly over any time...

Dry-Drink wrote: ↑Sat Jan 07, 2023 1:42 pm [The difference between the P/E of value and P/E of growth is called the "value spread".

So we can look at the ratio P/E Growth to P/E Value to estimate the value spread. If value had become increasingly popular, value stocks would have become more expensive relative to growth and the ratio would be smaller. Instead we still see the value spread at about the 90th percentile of historic values. Check out the below chart from Cliff Asness of AQR.

https://www.aqr.com/Insights/Perspectiv ... Not-Popped

Dave

I just retired at age 65 and have a more or less traditional roughly 55/45 nest egg IRA, along with a traditional pension. Thought that was conservative but the Bernstein statement threw me. (Am probably waiting to 70 to pick up SS, taking an additional 1% from the IRA in the meantime.) Wise counsel appreciated. Maybe clarification from Mr. Bernstein? Certainly depends what you consider “conservative”. This may help give some perspective. Stocks have a standard deviation of about 20%. Bonds might have a volatility of about 5%. That means that about 80-85% of the risk in your 55/45 portfolio is wrapped up in stocks. When making portfolio decisions, it’s valuable to think in terms of both potential % losses and potential absolute $ losses. D...