## Do you think about volatility drag on your portfolio?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Dirghatamas
Posts: 506
Joined: Fri Jan 01, 2016 6:18 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker I should be on my evening bike ride right now but partner didn't show up yet, so I will engage

This is actually a halfway decent article. I will quote the relevant piece

“Volatility drag” is one such conclusion. We found that the arithmetic and geometric means are related — both are equations using the same set of numbers — but the difference between them is in the definition, not from a force. Why, then, is this such a persistent myth? There are two seductive arguments made in its favor. One is very simple and appeals to our intuition, but contains a flawed assumption. The other stems from a slight misinterpretation of terms in a widely used mathematical model of prices."

I don't really have much more to say than that. There isn't a volatility drag. If the Boglehead Wiki has an article on that perhaps it should be worded more carefully. Anyone who went through middle school math (or maybe junior high) can do this type of math. If some one uses the wrong equation (addition instead of multiplication) then they deserve what they get. There has to be some some basic level of mathematical competency we expect investors to have!

Random Walker wrote:I found this link that explains what I'm talking about:

http://swanglobalinvestments.com/2016/0 ... is-a-drag/

Dave

This article is non sense. There is only one correct way of calculating returns over a period of time: (final value - initial value)/initial value. If two assets or portfolios deliver the same return (as in this formula), then they have the same return (as long as you are in accumulation and not withdrawing a large amount at the wrong time).

The illusion that there is a drag is just coming from using wrong math and wrong equations. Math is math.

There are some sensible reasons to hate volatility. If you are in withdrawal mode and need to withdraw a lot (say 30%) of your portfolio at a given instant, having a volatile portfolio could hurt you. That is obvious.

The parts I really do want to correct are 1) volatility by itself is not a drag 2) there isn't a re balancing bonus unless one can show hard back testing data. If we are fine on those two mathematical concepts, then we are good. I don't have anything further to say.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

Some people wade into the more mathematical conversations on this forum and it's so obvious that they are totally out of their depth.

MIpreRetirey
Posts: 632
Joined: Fri Sep 06, 2013 12:35 pm

### Re: Do you think about volatility drag on your portfolio?

Random.
Let me show you a tact of reason which I hope is useful and I can articulate it okay.

Identity properties:
The identity property of addition and subtraction says any number plus 0 = that number.
Also stated: if the difference between two numbers = 0, then the numbers are equal.

The identity property of multiplication says any number times 1 equals that number.
Also stated: if the product of two numbers = 1, then the numbers are the inverse of each.
Or, if the quotient of two numbers = 1, then the numbers are equal.

When you combine the gains of two or more years, you multiply.

So, to get an unchanged portfolio, if the 1st year gained 25% (5/4), and the next year gained -20% (4/5), the result is unchanged from the start.

Addition of +10 and -10% to equal 0 change is not what you'd do. Two year returns are multiplied. +10% is the ratio 11/10, -10% is the ratio 9/10.
So a +10% return followed by a -10% return would result in a total return of 11/10 * 9/10 = 99/100, or 9.9/10.

Simply saying use multiplication and a geometric mean (like CAGR is). And leave arithmetic mean for other things like statistics of things.
Your return is a geometric series of multiplications. ie: (1.10)(.95)(1.2)(1.1) = 1.3794. I don't care what the arithmetic average is when finding my returns. Don't even have to know what it is; don't care in this situation.

So, to go further. If I was looking to get an arithmetic average of 0, I could add +10%, -10%, +5%, -5%.
If I was looking to get a geometric return of 1, I could multiply +10%, -9.1%,+5%,-4.76%. (1.1 * 1/1.1 * 1.05 * 1/1.05).
Last edited by MIpreRetirey on Fri Mar 10, 2017 9:52 pm, edited 3 times in total.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Dirghatamas,
I agree there is only one way to calculate portfolio returns: at least that makes sense to me . But your example total return = final value - initial value / initial value misses something. The final values of two different portfolios can be different! And it can be due to differences in volatility.
Looking forward, if you can have either of two portfolios with the same expected mean return but different expected standard deviations, which would you choose? This is not only about the obvious stomach churning issue, but also about which portfolio is likely to have a higher end final value.
Assuming an individual chooses the portfolio with the lower SD, then the question arises how much more (if anything) is he willing to pay for the improvement on his portfolio. This is where the arguments over CCFs, AQR Style Premia, AQR Managed Futures come from. They are all expensive and essentially non correlated to the remainder of our typical Boglehead portfolios. Are the diversification benefits worth the cost? That is something we can all go round and round on. We might even argue about whether they provide benefit (I think they do). But I don't think it's arguable that a more efficient portfolio is preferable: same mean expected return with smaller SD (and tail protection) is better than same mean and higher SD. The reasonable price for improved efficiency is something very worthy of discussion.

Dave

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

345,
I'm wading into the math, and on the verge of being out of my league. Think I'm pretty much on target though.

Dave

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:345,
I'm wading into the math, and on the verge of being out of my league. Think I'm pretty much on target though.
You're doing fine. Not everyone is getting the point you're trying to make. I guess you're more patient than I am.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

MIPreretirey,
I understand that math. Do you agree that a more volatile portfolio with same average return will have lower terminal value than less volatile portfolio with same average return?

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

345,
I lost a bit in one post losing patience resulted in some very clear writing though

Dave

MIpreRetirey
Posts: 632
Joined: Fri Sep 06, 2013 12:35 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:MIPreretirey,
I understand that math. Do you agree that a more volatile portfolio with same average return will have lower terminal value than less volatile portfolio with same average return?

Dave

Nope. Because the arithmetic average is not the CAGR. (the geometric mean/average.) The volatility however large (unless you get a -100% in there) doesn't effect the CAGR by itself. Here's volatility with a net return of no change (1 * the original.)
5/4 * 25/10 * 80/10 ... * 10/80 * 10/25 * 4/5 = 1.

Some of those referenced links are not giving the up and up on what offsets a particular gain. Say +10%, -10% is not a net 0 change sequence. A net 0 change sequence would be +10%, -9.1% (just 1.1, 1/1.1 or .90909) (.90909 = -9.1%)

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Volatility does affect CAGR. It lowers CAGR. Prove it to yourself. Create two series of returns where both series has the same average annual return. Assign different standard deviations to the two series. Calculate the CAGR for the two series. You'll see that increased volatility, decreases CAGR. The greater the volatility, the bigger the difference between the weighted average annual portfolio return and the CAGR. Don't even consider average annual return as anything other than a forward looking expected weighted mean of portfolio components. Try different assigned SDs and you'll see the effect on terminal value.
In constructing a portfolio ex ante, we look at expected mean average annual returns for portfolio components, their individual SDs, and correlations. In evaluating the portfolio ex post, we look at the actual compounded CAGR and the portfolio SD. By minimizing the expected portfolio SD for a given expected return , we increase the chances of maximizing CAGR ex post relative to the returns of the portfolio components.

Dave

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

For those interested, I'd recommend reading Asset Allocation: Balancing Financial Risk 4th edition by Roger Gibson. Chapter 7 titled Diversification: TheThird Dimension. He gives good examples. He defines diversification effect as a beneficial reduction of portfolio volatility below the weighted average of the volatilities of the investments in the portfolio. When an asset with same expected return and same SD as portfolio but with less than +1 correlation is added, the expected return of the portfolio is unchanged, the expected SD is lowered, the end CAGR is likely higher, and the terminal value is likely higher.

Dave

Dirghatamas
Posts: 506
Joined: Fri Jan 01, 2016 6:18 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:Volatility does affect CAGR. It lowers CAGR. Prove it to yourself. Create two series of returns where both series has the same average annual return. Assign different standard deviations to the two series. Calculate the CAGR for the two series. You'll see that increased volatility, decreases CAGR. The greater the volatility, the bigger the difference between the weighted average annual portfolio return and the CAGR. Don't even consider average annual return as anything other than a forward looking expected weighted mean of portfolio components. Try different assigned SDs and you'll see the effect on terminal value.
In constructing a portfolio ex ante, we look at expected mean average annual returns for portfolio components, their individual SDs, and correlations. In evaluating the portfolio ex post, we look at the actual compounded CAGR and the portfolio SD. By minimizing the expected portfolio SD for a given expected return , we increase the chances of maximizing CAGR ex post relative to the returns of the portfolio components.

Dave

Dave

OK Dave I now understand that we were just talking past each other, rather than having an actual disagreement on a shared mental model (I finally grokked why we are arguing). Your mental model of the stock market (or any other asset) is a Random walk with known ARITHMETIC MEAN and standard deviation (place holder for volatility). You assume that these means and deviations are known ex-ante.

Clearly, in such a model, what you say above and in the OP is trivially obvious (and correct). The problem some of us are pointing out is we don't agree with this model. In my view stocks are NOT a Random Walk. They are partial ownership of the world's companies. So, ex-ante you have no idea for the holding period what the returns if any will be. The history of stocks is pretty short, the number of independent periods is small and the future may not at all look like the past.

As such, a model with arithmetic means and standard deviations etc. doesn't really mean much if used in the statistical sense. All you will get ex-post are the actual returns (you only get one shot). The returns will simply be (final value - initial value)/initial value. Based on that data, you can ex-post calculate the CAGR. What is gained by making it more complicated than that?

Now, if you had instead invested in a different asset class, which ended up with the same CAGR ex-post over your time period then by definition the CAGR of asset 1 & 2 are the same and so total returns are the same. That's really the only data you will have. We only live once and the future is unpredictable so we will only get one sequence of returns.

For that set of returns, again by definition, if the total returns are the same, then CAGR (geometric mean) will be the same. Yes indeed arithmetic means will turn out to have been different and the two asset classes may have (ex-post) shown very different variations and sequence of returns.

So, the question is how did volatility help you or hurt you. The answer is it didn't. If the CAGR (during your actual investing time) turned out the same, then returns were same, if CAGR was different, returns were different. What role did arithmetic mean or volatility play?

As for why the article you quote shows a different result: they simply compared two investments with different CAGR. Obviously in that case your returns are different..

Ari
Posts: 434
Joined: Sat May 23, 2015 6:59 am

### Re: Do you think about volatility drag on your portfolio?

I don't get it. Can someone answer why you're talking about the arithmetic mean? Is there any situation where this is a relevant number? Who would ever use the arithmetic mean for anything in evaluating a portfolio? It seems completely irrelevant.What is it good for?
All in, all the time.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Yes, we can't predict future. But normal distribution with mean and SD are not the worst approximations in the world. Using those approximations, we see that CAGR is hurt by volatility. So we should use that information to our advantage and construct portfolios with weakly, non, negatively correlated investments to minimize future portfolio volatility for a given targeted expected return.

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Ari,
When constructing a portfolio looking forward, it's useful to have a sense of each investment's addition to the portfolio: it's expected return, SD, correlations to other investments in the portfolio. The expected return used is a mean simple average return.
Yes, in hindsight, all that matters is the CAGR of our portfolio as a whole. But doesn't it make sense in portfolio design up front to want to make that future CAGR as close to the weighted average return of the portfolio components? It's going to be less by definition, but you can likely minimize how much less with portfolio design. Now these improvements all require increased cost beyond the cheapest TSM portfolios. Is improved portfolio efficiency worth it? That's where the arguments should be in my opinion. Once people appreciate the harmful effects of portfolio volatility, the tradeoffs of cost and improved portfolio efficiency are one issue. Another is what factors, asset classes, sources of return are worth it. These discussions all require thinking of an individual assets expected return, which is a simple average.

Dave

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

Ari wrote:I don't get it. Can someone answer why you're talking about the arithmetic mean? Is there any situation where this is a relevant number? Who would ever use the arithmetic mean for anything in evaluating a portfolio? It seems completely irrelevant.What is it good for?

The arithmetic mean is an estimate of the mean of the presumed statistical distribution from which successive annual returns are a sample. You use the same data to estimate the spread of that distribution, usually using the standard deviation as the statistic. The next step is to compound those returns in a statistical process to estimate the distribution of possible outcomes over time. When you do that one of the things you would compute is the geometric mean return for some time period, which is just an estimate from which you could compute the mean of the distribution of end point wealth, which is the result most people want to predict. One mathematical result of that calculation is that two portfolios that have the same arithmetic average annual return but different standard deviations of that annual return can have different average compound annual growth rates over time that are less for the more variable investment.

I think your point is that if you already know the geometric mean return, then why would you look at the arithmetic mean. And that is correct. You can approach the mathematical problem by looking at a set of geometric returns for lengths of time of interest and eventually extract the same calculations from what amounts to the same data.

Doing the math is just doing the math. I don't think it helps to introduce metaphors to these discussions because if you aren't thinking explicitly about the mathematics of the situation, metaphors simply introduce confusing references to what appear to be mysterious forces, malignant diseases of the portfolio, monsters under the bed, and so on. I object to those things whether they are "volatility drag," "sequence of returns risk," "diversifying over risk factors," "taxes mean the government owns part of your assets," or any of the other things of this ilk.

PS Random Walker and I are making the same statement exactly.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Dbr,
We're not exactly making same statement, you said it better! You said what I mean.

Dave

in_reality
Posts: 4189
Joined: Fri Jul 12, 2013 6:13 am

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:Volatility does affect CAGR. It lowers CAGR. Prove it to yourself.

Yes, it does (assuming equal average return). Still I neither think about volatility drag, nor care.

Random Walker wrote:Create two series of returns where both series has the same average annual return. Assign different standard deviations to the two series. Calculate the CAGR for the two series. You'll see that increased volatility, decreases CAGR. The greater the volatility, the bigger the difference between the weighted average annual portfolio return and the CAGR. Don't even consider average annual return as anything other than a forward looking expected weighted mean of portfolio components. Try different assigned SDs and you'll see the effect on terminal value.

I think that is pointless actually. I care about the CAGR and that's it.

In 10 years, if I believe emerging markets will have better returns (despite being more volatile) than alternative asset classes, I will invest in it.

Random Walker wrote:In constructing a portfolio ex ante, we look at expected mean average annual returns for portfolio components, their individual SDs, and correlations. In evaluating the portfolio ex post, we look at the actual compounded CAGR and the portfolio SD. By minimizing the expected portfolio SD for a given expected return , we increase the chances of maximizing CAGR ex post relative to the returns of the portfolio components.

I truly see no practical benefit in trying to predict average annual returns and selecting for lower deviation.

Here is a concrete example that I use in my asset allocation plan.

MSCI Emerging Markets have a 10 year expected real return of 7% and 23.3% volatility.
EAFE Equity has a 10 year expected real return of 5.4% and 17% volatility.
US Large has a 10 year expected real return of 0.7% and 14.4% volatility.

Real returns are based on expected growth, dividends, the effect of current valuations and forex effects.

What you are telling me to do is find an asset class that has the same average annual return as say Emerging Markets and choose on expected volatility, which I think is practically impossible.

I simply don't care what the average annual returns will be. I simply don't care what the volatility will be. If politics/the economy etc. go haywire, emerging markets are sure to see more volatility. Due to the way markets overreact and correct, the realized average annual return will be very different depending on how events go.

Despite that, I do believe that emerging markets will outperform based on their expected growth, dividends, the effect of current valuations and forex effects.

To artificially insist that I try to find a comparable asset class to emerging markets in terms of expected average annual returns, and then choose based on expected volatility isn't useful at all I think.

Yes, you are right! For the same average annual returns, the holding with lower volatility will have a higher CAGR. So what?

Can you give me a list of assets classes that have the same average annual returns so I can pick between them based on expected volatility?
[60% US _ 26% DEV _ 14% EM] | (-16% LC _ +8% MC _ +8% SC) | [47% FND/VAL _ 40% MKT _ 7% MOM _ 6% REIT] | (+/- 5% or *25% rebalancing bands)

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

in_reality wrote:
Can you give me a list of assets classes that have the same average annual returns so I can pick between them based on expected volatility?

That is not what people who want to worry about this do. What people who want to worry about this do is try to configure a whole portfolio that has the same or better average annual return but lower expected volatility. Generally the first step is to try to reduce the volatility by combining assets that have similar return and volatility but are not correlated.

However, you are correct that actually doing this in practice is an issue open to discussion. Those discussions are certainly available to be considered.

Since the OP asked if "I" think about volatility drag, the answer is I think about it as a mathematical problem but I don't try to engineer my portfolio to increase the compound growth using anything other than total market indices and generic bond funds.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

In_reality,
Just to take from your example. Which do you think is likely to be a better portfolio, 100% EM alone or 50% EM / 50% Int Small Value? Both are volatile high expected return asset classes. Correlation between the two is less than 1.

Dave

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:In_reality,
Just to take from your example. Which do you think is likely to be a better portfolio, 100% EM alone or 50% EM / 50% Int Small Value? Both are volatile high expected return asset classes. Correlation between the two is less than 1.

Dave

The essence of this portfolio being a combination of two asset classes that are not correlated rather than suggesting a choice of only one.

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

Like Ponce de Leon's search for the fountain of youth, investors look for the portfolio that has steady growth with minimal volatility. I think it is mostly a waste of time. I get portfolio theory and I get adding volatile non-correlating assets to a portfolio to (hopefully) increase returns a bit and at the same time reduce portfolio volatility. I have done the same things myself but none of it saved me from the 2008-2009 bear market and financial crisis. What worked brilliantly in 2000-2002 was a big fail in 2008-2009.

It reminds me of the brokerage newsletter highlighting the "ruler" stocks. That is the stocks that had 15% earnings growth that you could draw on a graph with a ruler. Just invest in those stocks and your worries were over. In retrospect, it was all baloney. The 15% growth was projected earnings growth and the 15% often never materialized in reality. Also, business fluctuates as does the rest of human life. Highs and lows, ups and downs. It was all fantasy. Learned later that the Coke CEO would call and berate stock analysts who dared suggest that Coke was not actually achieving that level of earnings growth. Jack Welch at GE timed asset sales to make earnings look smoother than they really were. A whole lot of financial engineering was going on out there.

In a similar fashion, we are cleverly engineering our portfolios to convince ourselves that our portfolios have higher returns and lower volatility than they do in real life. We are fooling ourselves that portfolio returns can be drawn upwards with a ruler. We just hate squiggly lines, we want graphs that look nice and neat.

All of this is inexact and a moving target, I know that I have tried. Sometimes all this portfolio tweaking seemed to add a benefit and sometimes it didn't. Not enough of a quant to do detailed spreadsheets, efficient frontiers, backtests, etc. Pretty much I have eyeballed things and thrown asset classes and sub-classes at the problem and crossed my fingers. It might help to research if there is a Patron Saint for portfolio volatility, that approach would likely work better than what I have done. The math and all of that are based on inexact numbers anyway.

Pretty much for me, two things help with volatility. I can use bonds in a portfolio to reduce volatility but at the cost of lower returns over time. Second, is time. Just wait it out and give the markets time to return to sanity. Really what this is all about boils down to this: how can I protect my portfolio from the effect of the extremes of human emotion? Why isn't investing 100% rational? With human emotion involved, investing can never be 100% rational, hence the excess volatility at times.
A fool and his money are good for business.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

Nedsaid,
Although I'm pondering all this ideal portfolio stuff, I also largely agree with you. Bogle quotes some general "enemy of good plan is the perfect plan". You also demonstrate that the fortitude to stick with your plan is more important than the plan itself.

Higher up in the thread though, Larry Swedroe made a great point. I think he was effectively saying that the major reason to build a robust portfolio diversified across sources of return is to create a plan the investor will stick with. That's the cake. My volatility drag issue is just icing at most.

Dave

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

Ari wrote:I don't get it. Can someone answer why you're talking about the arithmetic mean? Is there any situation where this is a relevant number? Who would ever use the arithmetic mean for anything in evaluating a portfolio? It seems completely irrelevant.What is it good for?
I said a similar thing
*3!4!/5! wrote:I can't think of a single reason for anyone to use average return instead of geometric return. It serves no purpose (other than to mislead - oh wait, maybe I just did think of a reason).

but I should reconsider. I'm sure this has been said earlier in various ways.

Geometric-expected-return is most relevant for investments "in series", i.e. in successive time periods 2016, 2017, 2018 etc. The returns are compounded (multiplied) and so geometric mean.

Arithmetic-expected-return is what is more natural for the math for investments "in parallel", i.e. an allocation to various investments for one time period, e.g. 20% of this 15% of that etc; the arithmetic-expected-return is a weighted sum of the separate returns.

Nevertheless, you want to see what happens in the long term, so you want to relate arithmetic-expected-return in a shorter time period to geometric-expected-return in that time period. That's what the math of this thread is about.

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:
Nevertheless, you want to see what happens in the long term, so you want to relate arithmetic-expected-return in a shorter time period to geometric-expected-return in that time period. That's what the math of this thread is about.

Yes, exactly so. That is why any discussion that you need one but not the other is already missing the point or maybe doesn't even have a point. You need what you need to do whatever it is you wanted to do.

Probably the answer to Random Walker's original question is that most of the people replying to this thread don't actually design their portfolios to try to increase geometric return by reducing portfolio volatility at the same expected return. More people probably try to increase geometric return by increasing expected return with more volatility. This would be, for example by holding more stocks and less bonds. It might be good if some people who do try what RW wants to do with his own investments actually post that they are doing it and how. It is unfortunate that so much of the discussion got side tracked.

To repeat what I do: No, I do not engage in designing a portfolio with the intent of reducing "volatility drag."

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

345,
Completely agree. Need to assign expected returns to each portfolio component when designing a portfolio. The portfolio expected return will be a weighted average.

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

DBR,
Oh ya, that's what my post was originally about. I almost forgot

Dave

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

For those interested, I just saw that William Bernstein discusses variance drag at the start of his book Rational Expectations, p.8-9.

Dave

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:For those interested, I just saw that William Bernstein discusses variance drag at the start of his book Rational Expectations, p.8-9.

Dave

I guess I think in narrative form because I am not much of a quant. Dr. Bernstein is a brilliant guy and I won't get into an argument with him. I am just not sure that variance drag or volatility drag really exists other than looking back and saying, "Gee whiz, if I had done thus and so back way back when, I could have done so much better." It almost seems like we are trying to be technical analysts ourselves while joining Burton Malkiel in our ridicule of it. Not many Bogleheads would admit to following technical analysis but it just seems we are doing something very similar. Instead of market timing, we are instead finding the "just right" mix of asset classes. We even use backtesting to design our efficient frontiers. Perhaps I am a bit of a pessimist after 30 years of trying, it seems like hope is triumphing over experience. Oh well, hope springs eternal.

It reminds me of a Larry Swedroe typo when he said that the Director of Research for his firm had a PhD in psychics. He meant physics but there might be more truth in his typo than he realized.
A fool and his money are good for business.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

nedsaid wrote:
Random Walker wrote:For those interested, I just saw that William Bernstein discusses variance drag at the start of his book Rational Expectations, p.8-9.

Dave

I guess I think in narrative form because I am not much of a quant. Dr. Bernstein is a brilliant guy and I won't get into an argument with him. I am just not sure that variance drag or volatility drag really exists other than looking back and saying, "Gee whiz, if I had done thus and so back way back when, I could have done so much better." It almost seems like we are trying to be technical analysts ourselves while joining Burton Malkiel in our ridicule of it. Not many Bogleheads would admit to following technical analysis but it just seems we are doing something very similar. Instead of market timing, we are instead finding the "just right" mix of asset classes. We even use backtesting to design our efficient frontiers. Perhaps I am a bit of a pessimist after 30 years of trying, it seems like hope is triumphing over experience. Oh well, hope springs eternal.

It reminds me of a Larry Swedroe typo when he said that the Director of Research for his firm had a PhD in psychics. He meant physics but there might be more truth in his typo than he realized.

This is a very major misreading of the situation. It's absolutely nothing like what you say. The concept of variance/volatility drag/drain can be presented as a piece of pure mathematics. This pure mathematics is absolutely correct, and there is nothing to dispute. People may object to the terminology, and they may argue the applicability of a math model to the real world. But the pure math itself is indisputable.

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:
nedsaid wrote:
Random Walker wrote:For those interested, I just saw that William Bernstein discusses variance drag at the start of his book Rational Expectations, p.8-9.

Dave

I guess I think in narrative form because I am not much of a quant. Dr. Bernstein is a brilliant guy and I won't get into an argument with him. I am just not sure that variance drag or volatility drag really exists other than looking back and saying, "Gee whiz, if I had done thus and so back way back when, I could have done so much better." It almost seems like we are trying to be technical analysts ourselves while joining Burton Malkiel in our ridicule of it. Not many Bogleheads would admit to following technical analysis but it just seems we are doing something very similar. Instead of market timing, we are instead finding the "just right" mix of asset classes. We even use backtesting to design our efficient frontiers. Perhaps I am a bit of a pessimist after 30 years of trying, it seems like hope is triumphing over experience. Oh well, hope springs eternal.

It reminds me of a Larry Swedroe typo when he said that the Director of Research for his firm had a PhD in psychics. He meant physics but there might be more truth in his typo than he realized.

This is a very major misreading of the situation. It's absolutely nothing like what you say. The concept of variance/volatility drag/drain can be presented as a piece of pure mathematics. This pure mathematics is absolutely correct, and there is nothing to dispute. People may object to the terminology, and they may argue the applicability of a math model to the real world. But the pure math itself is indisputable.

You are using numbers to express the effects of human emotion, that's all. Investing is not just a math problem to be solved, an awful lot of successful investing is emotional control. Again, I am not sure that portfolio drag is a real phenomenon. If volatility drag is such a certain thing, can't the quants tell us how to avoid it? My thesis is that they can't. Also, why don't the quant funds perform better than they do?
A fool and his money are good for business.

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:
This is a very major misreading of the situation. It's absolutely nothing like what you say. The concept of variance/volatility drag/drain can be presented as a piece of pure mathematics. This pure mathematics is absolutely correct, and there is nothing to dispute. People may object to the terminology, and they may argue the applicability of a math model to the real world. But the pure math itself is indisputable.

Indeed. And that is why the OP has a question about whether or not anyone is thinking about and applying this mathematical fact when it comes to configuring an actual portfolio. How the discussion wanders off to debating mathematics is just bizarre.

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

345,
I agree that the math is indisputable. I also think the applicability to the real world is indisputable too. The application is that looking forward, we should try to construct portfolios with enough non, weakly, negatively correlated factors as reasonable to try to minimize the difference between our Geo Mean and the weighted Average Simple Mean of portfolio components. The problem is that we are forced to trade increased costs for improved efficiency.
On the equity side, we take on some cost tilting to International. A little more cost going SV, more cost with AQR funds, and even more cost with alternative lending and reinsurance.

Dave

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

The actual question in this thread is whether or not people actually attempt to design portfolios that are on the efficient frontier. That means do they try to obtain the highest expected return possible at given risk or alternatively the lowest risk at a given expected return. Here the definition of risk is volatility, specifically the expected standard deviation of returns.

This is a classic problem in Modern Portfolio Theory originating with Markowitz.

Does anyone here use Financial Engines to guide their investment selection is another way to ask what people are doing.

kolea
Posts: 1234
Joined: Fri Jul 11, 2014 5:30 pm
Location: Maui and Columbia River Gorge

### Re: Do you think about volatility drag on your portfolio?

Ari wrote:I don't get it. Can someone answer why you're talking about the arithmetic mean? Is there any situation where this is a relevant number? Who would ever use the arithmetic mean for anything in evaluating a portfolio? It seems completely irrelevant.What is it good for?

The arithmetic mean is used a lot in MPT. It is necessary for calculating expected returns and efficiency frontiers.
Kolea (pron. ko-lay-uh). Golden plover.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

Random Walker wrote:345,
I agree that the math is indisputable. I also think the applicability to the real world is indisputable too. The application is that looking forward, we should try to construct portfolios with enough non, weakly, negatively correlated factors as reasonable to try to minimize the difference between our Geo Mean and the weighted Average Simple Mean of portfolio components. The problem is that we are forced to trade increased costs for improved efficiency.
On the equity side, we take on some cost tilting to International. A little more cost going SV, more cost with AQR funds, and even more cost with alternative lending and reinsurance.

Maybe I'm lazy. I just try to diversify ("the only free lunch") at low cost, so I'm happy with something resembling a three fund portfolio. (In my main retirement account there is a cheap SP500 fund and an expensive smallcap fund, so I skip the latter, though I fill that gap in another account.) I see the arguments for more complex strategies, but don't feel I have enough information to justify bothering with them (and again, in employer-linked retirement accounts, the possible choices are limited, so I'm happy to at least have a few low cost broad market funds - there's simply not the opportunity for something more complex even if I wanted to).

Random Walker
Posts: 1922
Joined: Fri Feb 23, 2007 8:21 pm

### Re: Do you think about volatility drag on your portfolio?

The cheapest and perhaps best diversified is high quality bonds. Most likely decreasing marginal benefit as one goes down my above list.

Dave

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:
Random Walker wrote:345,
I agree that the math is indisputable. I also think the applicability to the real world is indisputable too. The application is that looking forward, we should try to construct portfolios with enough non, weakly, negatively correlated factors as reasonable to try to minimize the difference between our Geo Mean and the weighted Average Simple Mean of portfolio components. The problem is that we are forced to trade increased costs for improved efficiency.
On the equity side, we take on some cost tilting to International. A little more cost going SV, more cost with AQR funds, and even more cost with alternative lending and reinsurance.

Maybe I'm lazy. I just try to diversify ("the only free lunch") at low cost, so I'm happy with something resembling a three fund portfolio. (In my main retirement account there is a cheap SP500 fund and an expensive smallcap fund, so I skip the latter, though I fill that gap in another account.) I see the arguments for more complex strategies, but don't feel I have enough information to justify bothering with them (and again, in employer-linked retirement accounts, the possible choices are limited, so I'm happy to at least have a few low cost broad market funds - there's simply not the opportunity for something more complex even if I wanted to).

This is odd, I have expressed skepticism over the phenomenon of volatility drag and yet I have attempted to smooth out my returns with non-correlating asset classes. I get told that I have it all wrong and yet that same poster says that a three fund portfolio is good enough. Am I missing something here? Why doesn't the true believer in portfolio drag attempt to do something about it?
A fool and his money are good for business.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

nedsaid wrote:
*3!4!/5! wrote:
Random Walker wrote:345,
I agree that the math is indisputable. I also think the applicability to the real world is indisputable too. The application is that looking forward, we should try to construct portfolios with enough non, weakly, negatively correlated factors as reasonable to try to minimize the difference between our Geo Mean and the weighted Average Simple Mean of portfolio components. The problem is that we are forced to trade increased costs for improved efficiency.
On the equity side, we take on some cost tilting to International. A little more cost going SV, more cost with AQR funds, and even more cost with alternative lending and reinsurance.

Maybe I'm lazy. I just try to diversify ("the only free lunch") at low cost, so I'm happy with something resembling a three fund portfolio. (In my main retirement account there is a cheap SP500 fund and an expensive smallcap fund, so I skip the latter, though I fill that gap in another account.) I see the arguments for more complex strategies, but don't feel I have enough information to justify bothering with them (and again, in employer-linked retirement accounts, the possible choices are limited, so I'm happy to at least have a few low cost broad market funds - there's simply not the opportunity for something more complex even if I wanted to).

This is odd, I have expressed skepticism over the phenomenon of volatility drag and yet I have attempted to smooth out my returns with non-correlating asset classes. I get told that I have it all wrong and yet that same poster says that a three fund portfolio is good enough. Am I missing something here? Why doesn't the true believer in portfolio drag attempt to do something about it?

How about you explain your understanding of volatility drag as a piece of pure mathematics.

Then explain how you can rebut pure mathematics with stuff like this:

nedsaid wrote: I am just not sure that variance drag or volatility drag really exists other than looking back and saying, "Gee whiz, if I had done thus and so back way back when, I could have done so much better."

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

^^ I should add that when it comes to volatility drag, diversification is doing something about it.

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:
nedsaid wrote:
*3!4!/5! wrote:
Random Walker wrote:345,
I agree that the math is indisputable. I also think the applicability to the real world is indisputable too. The application is that looking forward, we should try to construct portfolios with enough non, weakly, negatively correlated factors as reasonable to try to minimize the difference between our Geo Mean and the weighted Average Simple Mean of portfolio components. The problem is that we are forced to trade increased costs for improved efficiency.
On the equity side, we take on some cost tilting to International. A little more cost going SV, more cost with AQR funds, and even more cost with alternative lending and reinsurance.

Maybe I'm lazy. I just try to diversify ("the only free lunch") at low cost, so I'm happy with something resembling a three fund portfolio. (In my main retirement account there is a cheap SP500 fund and an expensive smallcap fund, so I skip the latter, though I fill that gap in another account.) I see the arguments for more complex strategies, but don't feel I have enough information to justify bothering with them (and again, in employer-linked retirement accounts, the possible choices are limited, so I'm happy to at least have a few low cost broad market funds - there's simply not the opportunity for something more complex even if I wanted to).

This is odd, I have expressed skepticism over the phenomenon of volatility drag and yet I have attempted to smooth out my returns with non-correlating asset classes. I get told that I have it all wrong and yet that same poster says that a three fund portfolio is good enough. Am I missing something here? Why doesn't the true believer in portfolio drag attempt to do something about it?

How about you explain your understanding of volatility drag as a piece of pure mathematics.

Then explain how you can rebut pure mathematics with stuff like this:

nedsaid wrote: I am just not sure that variance drag or volatility drag really exists other than looking back and saying, "Gee whiz, if I had done thus and so back way back when, I could have done so much better."

It sounds to me that you are saying that volatility drag is real but that you have chosen to do nothing about it. Perhaps as some have suggested you have decided that the implementation costs of trying to avoid it are too high. Or perhaps you don't want a more complex portfolio and what you have is good enough. Perhaps the effect is real but small enough that it isn't worth bothering with.

I guess the problem I have with all of this is that past performance does not translate to future results. Though we know how we could have avoided this effect in the past, we don't know what will work in the future. Is this something that is really actionable?

Sounds to me from your inaction that this isn't really a problem. If that is true, why is there a discussion?
A fool and his money are good for business.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

These threads just get crazy!

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

*3!4!/5! wrote:These threads just get crazy!

No, it isn't crazy. You argued very passionately about portfolio drag. I asked a very simple question. If this is such a big problem, why haven't you chosen to do anything about it?
A fool and his money are good for business.

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

dbr wrote:The actual question in this thread is whether or not people actually attempt to design portfolios that are on the efficient frontier. That means do they try to obtain the highest expected return possible at given risk or alternatively the lowest risk at a given expected return. Here the definition of risk is volatility, specifically the expected standard deviation of returns.

This is a classic problem in Modern Portfolio Theory originating with Markowitz.

Does anyone here use Financial Engines to guide their investment selection is another way to ask what people are doing.

My answer is yes, I have attempted use efficient frontiers to design my portfolio. Yes, I have attempted to use Modern Portfolio Theory. And yes, I used Financial Engines in the past but I was dissatisfied with the results, it wanted a US Large Cap heavy portfolio with much less emphasis upon International Stocks and Mid-Cap/Small Cap than what I wanted.

I was heavily influenced by Paul Merriman and his company and his presentation of academic research. I found the case for small/value tilting to be quite persuasive.

I also have used Retirement Target Date and Target Risk portfolios as a template for my own portfolio. It is a way of using the best thinking of experts at mutual fund companies to construct my own portfolios.

I get the concepts and have used them myself, I am not sure that I actually got the results I was hoping for.
A fool and his money are good for business.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

Reposting:
viewtopic.php?p=3278479#p3278479
*3!4!/5! wrote:^^ I should add that when it comes to volatility drag, diversification is doing something about it.

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

dbr wrote:
Doing the math is just doing the math. I don't think it helps to introduce metaphors to these discussions because if you aren't thinking explicitly about the mathematics of the situation, metaphors simply introduce confusing references to what appear to be mysterious forces, malignant diseases of the portfolio, monsters under the bed, and so on. I object to those things whether they are "volatility drag," "sequence of returns risk," "diversifying over risk factors," "taxes mean the government owns part of your assets," or any of the other things of this ilk.

Math can be very, very dry. It is also very important in investing, but math isn't the whole thing. Human nature, human behavior, and human emotion are very important to understanding how all this works. Investing can't just be reduced to formulas.

What I am attempting and apparently failing to do is to make arguments in narrative form. Tell a story and give examples to illustrate my points. We could make this all a dry recitation of facts and mathematical formulas but no one would read this forum but mathematicians, statisticians, actuaries, and maybe engineers. It would be as interesting as reading the New York City phone book from cover to cover.

Part of what I am doing is shorthand thinking. I don't want my posts to run many paragraphs. I am trying but apparently failing to boil things to their essence and explain them in a fun and enjoyable way. This isn't appealing to mysterious forces or monsters under the bed. Quite frankly, I really resent this and if all you want are incredibly dry discussions, I will just butt the heck out.

My gosh, you want this to be something that people will actually read and maybe have a little fun along the way. Apparently, this is all frowned upon here.
A fool and his money are good for business.

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

I suspect the essence of the problem is that actually computing an asset allocation nearer to or on the available efficient frontier is hard to do given

1. difficulty in estimating mean, volatility, and correlation of returns
2. counting on same to be stable enough over time to reasonably implement
3. difficulty to find volatile investments that are sufficiently uncorrelated to get stable solutions in MVO
4. sensitivity to assumptions that are not valid, such as the assumed distribution of returns not being well understood (aka not normal or even worse than merely being not normal)

Maybe someone else has a better insight into the present status of this area of MPT as a practical portfolio management tool.

dbr
Posts: 22901
Joined: Sun Mar 04, 2007 9:50 am

### Re: Do you think about volatility drag on your portfolio?

nedsaid wrote:
dbr wrote:
Doing the math is just doing the math. I don't think it helps to introduce metaphors to these discussions because if you aren't thinking explicitly about the mathematics of the situation, metaphors simply introduce confusing references to what appear to be mysterious forces, malignant diseases of the portfolio, monsters under the bed, and so on. I object to those things whether they are "volatility drag," "sequence of returns risk," "diversifying over risk factors," "taxes mean the government owns part of your assets," or any of the other things of this ilk.

Math can be very, very dry. It is also very important in investing, but math isn't the whole thing. Human nature, human behavior, and human emotion are very important to understanding how all this works. Investing can't just be reduced to formulas.

What I am attempting and apparently failing to do is to make arguments in narrative form. Tell a story and give examples to illustrate my points. We could make this all a dry recitation of facts and mathematical formulas but no one would read this forum but mathematicians, statisticians, actuaries, and maybe engineers. It would be as interesting as reading the New York City phone book from cover to cover.

Part of what I am doing is shorthand thinking. I don't want my posts to run many paragraphs. I am trying but apparently failing to boil things to their essence and explain them in a fun and enjoyable way. This isn't appealing to mysterious forces or monsters under the bed. Quite frankly, I really resent this and if all you want are incredibly dry discussions, I will just butt the heck out.

My gosh, you want this to be something that people will actually read and maybe have a little fun along the way. Apparently, this is all frowned upon here.

My comment was not intended to be personal. I apologize if that is the impression given.

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

### Re: Do you think about volatility drag on your portfolio?

dbr wrote:I suspect the essence of the problem is that actually computing an asset allocation nearer to or on the available efficient frontier is hard to do given

1. difficulty in estimating mean, volatility, and correlation of returns
2. counting on same to be stable enough over time to reasonably implement
3. difficulty to find volatile investments that are sufficiently uncorrelated to get stable solutions in MVO
4. sensitivity to assumptions that are not valid, such as the assumed distribution of returns not being well understood (aka not normal or even worse than merely being not normal)

Maybe someone else has a better insight into the present status of this area of MPT as a practical portfolio management tool.

Right. It seems futile to crunch through a bunch of data (especially for a retail investor with limited options to even act upon their dubious conclusions). I would never consider doing that. But some general investing principles survive all the difficulties you list, and the obvious one that jumps out in the context of volatility drag is simply:

"Diversify!"

nedsaid
Posts: 8289
Joined: Fri Nov 23, 2012 12:33 pm

### Re: Do you think about volatility drag on your portfolio?

dbr wrote:I suspect the essence of the problem is that actually computing an asset allocation nearer to or on the available efficient frontier is hard to do given

1. difficulty in estimating mean, volatility, and correlation of returns
2. counting on same to be stable enough over time to reasonably implement
3. difficulty to find volatile investments that are sufficiently uncorrelated to get stable solutions in MVO
4. sensitivity to assumptions that are not valid, such as the assumed distribution of returns not being well understood (aka not normal or even worse than merely being not normal)

Maybe someone else has a better insight into the present status of this area of MPT as a practical portfolio management tool.

What I would say is that past performance of asset classes gives us a good idea of how they might act in the future. Just because calculating an asset allocation that optimizes risk and return in the future is impossible doesn't mean that we shouldn't try. An educated guess based on past history and future projections is a whole lot better than nothing at all. It is all imperfect but still very useful.

This is why I have eyeballed a lot of things and not tried to get too fancy about this. It is all imprecise anyway. I do think an investor can improve his or her odds but there has to be an understanding that the tools are not perfect.

I don't want anyone to interpret that I have just dismissed all the arguments made here. My gosh, we all want to boost returns a bit and reduce volatility. If I thought this was all bunk, why would I have made all the effort to small/value tilt my portfolio, internationally diversify, and add REITs? Why would I go to Merriman seminars, read Larry Swedroe, and bone up on Modern Portfolio Theory. I must have thought there was something to all of that. Though I have probably not expressed myself too well, I want reduce volatility in my portfolio too. I don't want to see the wild swings in my account values.

Pretty much, I have made the effort and wondered if I really got the results I wanted.
A fool and his money are good for business.