Larry Swedroe: Managing Risk With Factors

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
TD2626
Posts: 609
Joined: Thu Mar 16, 2017 3:40 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by TD2626 » Wed Nov 22, 2017 11:33 am

Random Walker wrote:
Wed Nov 22, 2017 12:01 am
As I’ve written many times in the past, costs are certain and potential benefits are only just that, potential. So it’s hard to go wrong with a low cost tax efficient TSM approach. I would consider that the benchmark for comparison. But take a look at my thread on volatility drag, or Larry’s short essay Efficient Diversification In A 3 Factor World, or the central chart in Gibson’s Asset Allocation. Volatility inflicts a real cost on a portfolio. (And this ignores perhaps the bigger issue of investor behavior in the face of a grueling equity bear market.). Whether one looks at a more efficient portfolio as one that lessens portfolio standard deviation and increases Sharpe ratio, one that lessens maximal drawdown, one that brings geometric return closer to average annual return, the conclusion is the same. A more efficient portfolio is worth some increased cost compared to the rock bottom cost of TSM/TBM. Is it possible the advisor/DFA/AQR/factor route increases the cost too much? The answer quite possibly is yes. But after a lot of ongoing thought, I chose that route because I saw the number and size of potential benefits start to look large compared to the certain costs.
Unfortunately, as I have learned, there is no way to directly compare the two paths. The sample period would be too short as well. They are apples and oranges. My overall stock/bond split is much different than I would have likely put together on my own, I doubt I would have ever used alternatives on my own, and I’m sure I would have made tweaks in my all VG portfolio over time as well. I think my behavior as an investor is pretty solid, but I nonetheless have an extra level of protection from myself with an advisor.
Currently we have had an 8-9 year bull market. We all look like investing superstars. A 100% equity growthy TSM portfolio looks like the MVP. Let’s see what the conversations look like when we have a bear like 1972-4 or worse yet Japan for a generation. There are many potential alternative histories, but only one will play out. Investing is about putting the odds of success in our favor. Starting with rock bottom costs is about as strong a start as one could ask for. Betting on the US economic engine is certainly pretty good as well. But academic research has shown we can diversify way beyond that in multiple dimensions. That diversification is worth something.

Dave
Very good points - I like the point about how a 3-fund portfolio should be seen as a benchmark for performance comparison. I like to think of it as a benchmark portfolio. My thinking is this, though - if it's so hard to beat the benchmark portfolio, why not own the benchmark portfolio?

Also - good point about how one should consider not just how things did turn out, but the full spectrum of what was possible. Not confusing strategy with outcome is important. If one takes steps to prevent a "Japan scenario" (e.g. diversifying away from the home country) and the home country doesn't experience a Japan-scenario meltdown, of course the added expense and volatility drag of currency risk would likely pull down returns. That's a cost of the protection. Insurance isn't a bad deal just because you didn't use it.

Finally, I think that it is best to note that there can be different portfolios for different people. If someone is particularly interested in maximizing portfolio efficiency, and has the knowledge, experience, and wherewithal to invest in a way that aims to do that, maybe that's reasonable for them. Most people probably can't manage this sort of complex portfolio - so TSM is likely better for them. The improvements of a complex portfolio are likely at the margins and slight (and aren't guaranteed... the only thing guaranteed by a complex strategy is more paperwork, hassle, and expense). But there can be "more than one road to Dublin" as Taylor often has said and as I feel Rick was echoing earlier in this thread.

It's reasonable to think that Dave likely has one of the more efficient portfolios around, and in some ways that is enviable. However, in other ways it begs the question -- why is having the most efficient portfolio possible the overriding aim? Yes, having a portfolio that has the highest return per unit risk (with risk measured by standard deviation) is a good goal. However, maybe there are more important things to life, and settling for a reasonably efficient portfolio (like a 3-fund portfolio) and enjoying hobbies or new-found spare time could be better. (Of course, if investing is intellectually stimulating and is a hobby in and of itself, this could be different.)

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Wed Nov 22, 2017 11:53 am

Triceratop,
Thanks for clarifying my point. Correlations of all risky assets do strongly tend to 1 in bad bears. Bonds, which are generally uncorrelated to risky assets, have a thankful tendency for correlations to go negative at those times. I think the value premium is likely at least in part compensation for doing really badly at those times all risky assets do badly together. Given the persistence of human behavior, I think Time Series Momentum has good potential to be a bit of a saving grace in bears. Can’t see any reason for reinsurance to correlate with equity markets. I hope the short duration nature of alternative lending increases the relative safety of this investment in bad equity markets. The AQR Style Premia Fund has 4 styles applied to I think 4 asset classes. Everything within that fund is uncorrelated with everything else in the fund and the fund is uncorrelated with everything outside it. Moreover it’s market neutral. So I think it’s reasonable to expect it to not also fall into the correlations go to 1 hole. I think the Stone Ridge All Asset Variance Risk Premium Fund can take it on the chin when markets tank, but that’s exactly what one would expect when selling insurance against markets tanking. Like reinsurance, the premiums should be highest after exiting the insurance disaster.
The biggest point is the one you alluded to. Increasing the risks with uncorrelated factors on the equity side allows one to increase the allocation to safe bonds.
I think how one view the alternatives in the portfolio depends on whether they take from equities or bonds to create the position. Either way, portfolio efficiency is increased at increased cost.

Dave

User avatar
digarei
Posts: 759
Joined: Sat Jul 05, 2014 1:41 am
Location: Sacramento
Contact:

Re: Larry Swedroe: Managing Risk With Factors

Post by digarei » Wed Nov 22, 2017 12:13 pm

willthrill81 wrote:
Wed Nov 22, 2017 11:15 am
digarei wrote:
Tue Nov 21, 2017 3:26 am
Random Walker wrote:
Mon Nov 20, 2017 11:05 pm
Dead Man Walking,
Like all the factors, Cross Sectional Momentum is defined as a long-short portfolio to isolate the factor and exclude the market factor. CS Momentum is RELATIVE momentum. Within an asset class, at defined time intervals, one could simply look back at the last time period and buy the top 30% performers and sell the bottom 30% performers. If this occurs at defined intervals, there is no effort to time the market or time the performance of any individual security. Note that these sales and purchases occur whether the market is up or down, because what is being evaluated is relative momentum. The manager could be buying stocks that went down the least and selling what went down the most, or he could be buying what went up most and selling what went up least.
Even time series momentum can be passive and formulaic. Time series momentum is absolute momentum: buy what has gone up in past and sell what has gone down. This therefore is not market neutral and is trend following. Thus this perhaps can be considered a passive formulaic approach to market timing? In the equities markets, TS Mom will result in increased equity allocation when markets have been up and decreased equity allocation when equities have been down. But even this strategy is agnostic to the specific stocks and agnostic to the market direction; it will formulaicly buy and sell as prescribed.
I’m pretty much all in on factors. I find it fascinating that one of the factors I intuitively have the least confidence in TS Mom (100% behavioral), has some of the strongest data behind it.

Dave

  • This seems to be the inverse of rebalancing. But not market timing? Longtime Bogleheads perhaps employ a different use of the phrase market timing than I am accustomed to:

What is 'Market Timing'

Market timing is the act of moving in and out of the market or switching between asset classes based on using predictive methods such as technical indicators or economic data.

Because it is extremely difficult to predict the future direction of the stock market, investors who try to time the market, especially mutual fund investors, tend to underperform investors who remain invested.

Upon investigation of the major marketing timing strategies (most investors who practice market timing seem to do so based on feelings), it seems that they tend to underperform buy-and-hold during bull markets and outperform during bear markets and, potentially, over long-term periods.

For instance, if one employed a 200 day moving average strategy where you invested in VTSMX (Vanguard Total Stock Market) when it was above the 200 DMA and switched to VBMFX (Vanguard Total Bond Market) otherwise, your return from 2009 until today would 14.42%, compared to 15.17% for buy-and-hold of VTSMX. But if we change the period of analysis to 1993 so that two major bear markets are included in the analysis, the timing model had a return of 11.76%, whereas buy-and-hold returned 9.51%. On top of that, the timing model had far lower volatility; the worst year for it was -5.89% with a max drawdown of -17.57%, compared to -37.04% and -50.89%, respectively, for buy-and-hold. Many were recommending this very simple approach 50 years ago, and it seems that those who practiced it rigidly did very well.

Mainly due to the seeming ability of market timing strategies to reduce downside volatility, it seems that they could hold substantial promise for retirees in the withdrawal phase. The reduction of downside risk could go a very long way to dramatically reducing sequence of returns risk, which is what has historically held safe withdrawal rates significantly below 'average' withdrawal rates.
I guess the main point in my reply to Random Walker is that with all of the buying and selling requisite to achieving certain factors, it is not as was described—it’s market timing of a sort, and not passive. I wasn’t suggesting that a market timing approach couldn’t be used successfully in some instances, by some investors, in some periods (I haven’t done the homework on this) but let’s agree that it is an active approach. I think it’s fair to categorize rebalancing in the same way.

Knowing someone who used a moving average strategy in the 1980s (evidently, it worked until it didn’t) makes me reluctant to endorse such methods. Who am I competing with when pursuing this alpha? What are the trading costs? What is my time worth? Can I keep this up for 10 years (nevermind, 30 yrs. while in retirement)? Would I want to if I could?

Fortunately, there are other ways to manage sequence of returns risk.
Connect with Bogleheads in Northern California! Click the link under my user info/avatar.

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Wed Nov 22, 2017 1:19 pm

I would argue that accessing the CS momtum or TS Momentum premia is not alpha. They are known premia that are readily investable in a passive formulaic fashion. As Larry has written, what was once alpha has been turned into a beta.

Dave

vencat
Posts: 207
Joined: Thu Sep 10, 2009 6:30 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by vencat » Wed Nov 22, 2017 1:51 pm

Dave,
Can you give me a comparative alt portfolio (including MOM, Reinsurance etc.) to a 60:40 three fund one. I really want to follow this prospectively.
The alt portfolio has good 'science' but I'm skeptical of costs, implementation and performance in the next bear market and and I'm not even going in to tax considerations.
Unfortunately with investing, I believe, one has to embrace and accept 'fat tails' with lots of patience, discipline and a good dose of bonds.

User avatar
willthrill81
Posts: 4161
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: Managing Risk With Factors

Post by willthrill81 » Wed Nov 22, 2017 2:17 pm

digarei wrote:
Wed Nov 22, 2017 12:13 pm
I guess the main point in my reply to Random Walker is that with all of the buying and selling requisite to achieving certain factors, it is not as was described—it’s market timing of a sort, and not passive. I wasn’t suggesting that a market timing approach couldn’t be used successfully in some instances, by some investors, in some periods (I haven’t done the homework on this) but let’s agree that it is an active approach. I think it’s fair to categorize rebalancing in the same way.
The 'most' passive of all portfolios is a global cap-weighted portfolio that is bought and held. Anything else is taking on some degree of 'activeness'.
digarei wrote:
Wed Nov 22, 2017 12:13 pm
Knowing someone who used a moving average strategy in the 1980s (evidently, it worked until it didn’t) makes me reluctant to endorse such methods.
Well the data are very clear that at least this specific strategy has worked well for the last 25 years. The biggest 'catch' is that those implementing it are likely to underperform a buy-and-hold approach when times are good (e.g. 2009-Current). Many would not have the fortitude to deal with this for a decade or more between cycles. Of course, the same could be said of people practicing buy-and-hold who panic in bear markets. Either way, you need determination to stick with your strategy.
digarei wrote:
Wed Nov 22, 2017 12:13 pm
What are the trading costs? What is my time worth? Can I keep this up for 10 years (nevermind, 30 yrs. while in retirement)? Would I want to if I could?
Trading costs are far less for most of us these days than they once were. I can trade all of the funds I have access to for free.

Checking your assets' performance once a month for a few minutes at most isn't a major time commitment. But well into retirement, you might want someone else to do this for you. Paul Merriman has used market timing successfully (he uses it primarily to minimize downside risk, not as an attempt to outperform market returns) since the 1980s but now has his son-in-law, a financial adviser, take care of this for him now. He just doesn't want to do it anymore.
digarei wrote:
Wed Nov 22, 2017 12:13 pm
Fortunately, there are other ways to manage sequence of returns risk.
Very true. There are many roads to Dublin.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
digarei
Posts: 759
Joined: Sat Jul 05, 2014 1:41 am
Location: Sacramento
Contact:

Re: Larry Swedroe: Managing Risk With Factors

Post by digarei » Wed Nov 22, 2017 4:09 pm

willthrill81 wrote:
Wed Nov 22, 2017 2:17 pm

Checking your assets' performance once a month for a few minutes at most isn't a major time commitment. But well into retirement, you might want someone else to do this for you. Paul Merriman has used market timing successfully (he uses it primarily to minimize downside risk, not as an attempt to outperform market returns) since the 1980s but now has his son-in-law, a financial adviser, take care of this for him now. He just doesn't want to do it anymore.
The market timing approach that I referenced earlier required the input of market data on a daily basis (from the WSJ) into a computer program, which rendered a verdict: SELL or BUY. When the signal changed, the investor phoned in a transfer order to the brokerage, moving sums between an equity fund and a money market account. As I recall, sometimes there were no trades for a month or more, and at other times, orders were made 2 or more times in a single week.

Regardless, it was work, consuming 15-30 minutes every day the NYSE was open for business.

You may surmise without fear of contradiction that I was a party to this drudgery. Wildly successful for several years, the last year this approach resulted in big losses. Finally, this effort was abandoned when the brokerage implemented a new frequent-trading policy that restricted the number of exchanges in any month. In the end, cumulative returns trailed the S&P.
Connect with Bogleheads in Northern California! Click the link under my user info/avatar.

User avatar
willthrill81
Posts: 4161
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: Managing Risk With Factors

Post by willthrill81 » Wed Nov 22, 2017 4:13 pm

digarei wrote:
Wed Nov 22, 2017 4:09 pm
willthrill81 wrote:
Wed Nov 22, 2017 2:17 pm

Checking your assets' performance once a month for a few minutes at most isn't a major time commitment. But well into retirement, you might want someone else to do this for you. Paul Merriman has used market timing successfully (he uses it primarily to minimize downside risk, not as an attempt to outperform market returns) since the 1980s but now has his son-in-law, a financial adviser, take care of this for him now. He just doesn't want to do it anymore.
The market timing approach that I referenced earlier required the input of market data on a daily basis (from the WSJ) into a computer program, which rendered a verdict: SELL or BUY. When the signal changed, the investor phoned in a transfer order to the brokerage, moving sums between an equity fund and a money market account. As I recall, sometimes there were no trades for a month or more, and at other times, orders were made 2 or more times in a single week.

Regardless, it was work, consuming 15-30 minutes every day the NYSE was open for business.

You may surmise without fear of contradiction that I was a party to this drudgery. Wildly successful for several years, the last year this approach resulted in big losses. Finally, this effort was abandoned when the brokerage implemented a new frequent-trading policy that restricted the number of exchanges in any month. In the end, cumulative returns trailed the S&P.
Yes, daily trading like that is for the birds unless it is consistently very successful.

The analysis I referenced earlier limited trades to one per month.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Larry Swedroe: Managing Risk With Factors

Post by in_reality » Wed Nov 22, 2017 9:55 pm

Random Walker wrote:
Wed Nov 22, 2017 12:01 am
But academic research has shown we can diversify way beyond that in multiple dimensions. That diversification is worth something.
Investing in part is emotional and one researcher writing in the Yale tradition of scholarly work on emotional intelligence writes:
In many academic and clinical endeavors, “confusion of tongues” hinders understanding and meaningful dialog. The 2016 Oxford dictionary has given the word “post-truth” term of the year status because of the current fragmentation of meaning, if not gross bias, in modern discourse. Because of current trends toward defining and viewing terms differently, the euphemistic treadmill and rebranding that may cause confusion and misunderstanding are (best) recognized, minimized, and avoided, when possible.
I do believe that people's understanding of factor investing would be enhanced by taking such an approach aimed at minimizing and avoiding confusion in relation to the use of "diversification" which we know is an emotionally attractive word.

I truly do wish factor investing was logically consistent, and didn't try co-opt the notion of diversification.

Take for instance the claims that the market portfolio only has exposure to beta, and that the market portfolio has no net value exposure, and that by adding value exposure by shorting growth stocks you will add diversification.

In fact, we can divide stocks into high and low beta, short the high beta stocks and end up with the BAB factor.

So to me a logical analysis would explain returns in terms of beta, size, style, investment, and profitability.

And a market portfolio would be the most diversified because it contains everything -- high beta, low beta, small size, large size, value style, growth style, high investment, low investment, high profitability and low profitability.

It's inconsistent to say we have beta exposure by holding the total market when we know that means both high and low beta stocks.

Yet, nobody says someone lacks exposure to beta because they don't have a net low beta exposure.

Nobody would say that holding VT (Vanguard Total World Stock) gives you no US exposure because you are holding the US at market cap weight.

Yet, if you hold VT (Vanguard Total World Stock), then suddenly you lack low-beta, small size, value style, low investment, high profitibablity exposure.

I personally find the “post-truth” use of this notion of diversification to be logical inconsistent.

Sure, targeting low beta, low investment, and small size is different that only targeting small size, but is additionally excluding high-beta and large size really diversifying you more than only excluding high investment?

Will excluding both emerging markets and developed international leave you more diversified? Factor adherents say it will.

I challenge the logic behind that and protest “post-truth" uses of terminology to associate tilting with the emotionally positive concept of diversification.

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Wed Nov 22, 2017 11:14 pm

In Reality,
Let me give a try at clarifying my use of diversification benefit. Some people say that adding 1 stock to a portfolio of 1000 stocks increases diversification. That logic results in the philosophy that TSM is most diversified. But none of us is concerned with the world market of all financial assets. Rather we are concerned with individual portfolios. When we look at a typical chart of return on the Y axis and SD on the X axis, we are interested in getting as close to the northwest corner efficient frontier as possible. A new portfolio addition provides a diversification benefit when it likely moves the portfolio closer to the northwest corner. The only way a new portfolio addition will do that is if it provides some expected return, it has low/non/negative correlation with other portfolio components, and has some volatility.
Another way to look at diversification benefit is the Larry Swedroe view. If an additional investment to a portfolio can keep the mean expected return constant but narrow the dispersion of returns and lessen the tails, then that investment has provided a meaningful diversification benefit to the portfolio.
It’s all modern portfolio theory. Diversification increases return per unit risk, Sharpe ratio. From that perspective, adding another large cap stock to TSM doesnt really add any diversification benefit to the portfolio. Adding a volatile, uncorrelated, high expected return, investment does provide diversification benefit to the portfolio. It’s about the behavior of individual portfolios.

Dave

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Larry Swedroe: Managing Risk With Factors

Post by in_reality » Thu Nov 23, 2017 1:36 am

Random Walker wrote:
Wed Nov 22, 2017 11:14 pm
Adding a volatile, uncorrelated, high expected return, investment does provide diversification benefit to the portfolio. It’s about the behavior of individual portfolios.
Hi Dave,

I'm sorry. I do hear and understand your meaning, but don't see the logic behind it.

If I were to take only a factor approach and hold small size, value style, low-beta, low-investment and high profitability, my expected returns are expected to be higher (by factor believers). That is true correct?

And yet you assert that large size, growth style, high-beta, high-investment, and low profitability are uncorrelated to those other factors (above).

So why isn't adding the volatile, uncorrelated, high expected returns of large size, growth style, high-beta, high-investment, and low profitability diversifying?

In other words, if high-beta and low-beta are uncorrelated, then how does only holding low-beta stocks add diversification? Clearly the factor position is that low-beta stocks will outperform. Yet clearly too, you aren't outperforming due to the diversification benefit of holding uncorrelated assets (because you aren't holding the uncorrelated high-beta stocks).

Are you asserting that value stocks will only outperform if held in a portfolio that includes growth? That is what your position would seem to imply.

Call_Me_Op
Posts: 6815
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry Swedroe: Managing Risk With Factors

Post by Call_Me_Op » Thu Nov 23, 2017 7:12 am

Random Walker wrote:
Wed Nov 22, 2017 9:55 am
Taylor,
Totally agree. Not only are high quality bonds the best protection, they are the cheapest as well. Everything else I discuss is decreasing marginal benefit and increasing marginal cost. This is especially true since the allocations are all quite small.

Dave
Yes, but we are not comparing stocks with bonds here. Clearly, there are benefits of having bonds for downside protection. We are discussing possible benefits of "diversifying" the stock portion of the portfolio by weighting factors higher than their representation in total market index.

I honestly think folks worry too much about their allocation within the equity bucket. The most important decision, and one which is different for everyone, is the appropriate ratio of stocks to quality bonds. Today, you can invest in stocks cheaply, whether you use total market index of factor indexes.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 9:13 am

Call me op,
It’s all linked together though. How much one tilts affects the choice of bond allocation. A person may decide on a reasonable expected return from a portfolio. He can then attempt to generate a portfolio with that expected return. If he does so with a TSM type portfolio, he will need more market beta risk and fewer safe bonds. If he decides to tilt the equity portion of the portfolio to higher expected return asset classes, then he can have a higher safe bond allocation. The expected return of the portfolio is greater, the expected volatility less, and the risks spread more evenly among weakly correlated independent factors. Term and credit are factors as well.

Dave

betablocker
Posts: 366
Joined: Mon Jan 11, 2016 1:26 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by betablocker » Thu Nov 23, 2017 9:27 am

Good clarification. You are correct to say that adding high beta, growth stocks provides diversification of outcomes. It’s just that the long term expected outcomes are lower. The goal of a factor investor is not diversification alone. It’s to find as many persistent and unique sources of risk with high expected returns and put them together. Low returning strategies provide diversification without return or deworsification.
in_reality wrote:
Thu Nov 23, 2017 1:36 am
Random Walker wrote:
Wed Nov 22, 2017 11:14 pm
Adding a volatile, uncorrelated, high expected return, investment does provide diversification benefit to the portfolio. It’s about the behavior of individual portfolios.
Hi Dave,

I'm sorry. I do hear and understand your meaning, but don't see the logic behind it.

If I were to take only a factor approach and hold small size, value style, low-beta, low-investment and high profitability, my expected returns are expected to be higher (by factor believers). That is true correct?

And yet you assert that large size, growth style, high-beta, high-investment, and low profitability are uncorrelated to those other factors (above).

So why isn't adding the volatile, uncorrelated, high expected returns of large size, growth style, high-beta, high-investment, and low profitability diversifying?

In other words, if high-beta and low-beta are uncorrelated, then how does only holding low-beta stocks add diversification? Clearly the factor position is that low-beta stocks will outperform. Yet clearly too, you aren't outperforming due to the diversification benefit of holding uncorrelated assets (because you aren't holding the uncorrelated high-beta stocks).

Are you asserting that value stocks will only outperform if held in a portfolio that includes growth? That is what your position would seem to imply.

Call_Me_Op
Posts: 6815
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry Swedroe: Managing Risk With Factors

Post by Call_Me_Op » Thu Nov 23, 2017 9:42 am

Random Walker wrote:
Thu Nov 23, 2017 9:13 am
Call me op,
It’s all linked together though. How much one tilts affects the choice of bond allocation. A person may decide on a reasonable expected return from a portfolio. He can then attempt to generate a portfolio with that expected return. If he does so with a TSM type portfolio, he will need more market beta risk and fewer safe bonds. If he decides to tilt the equity portion of the portfolio to higher expected return asset classes, then he can have a higher safe bond allocation. The expected return of the portfolio is greater, the expected volatility less, and the risks spread more evenly among weakly correlated independent factors. Term and credit are factors as well.

Dave
Indeed - but "expected returns" vary with who is doing the expecting. For full disclosure, I am somewhat of a tilter. I in fact hold a "Larry-like" portfolio. However, I do not have an "expected return." What I do have is an "expected draw-down", which is 50% of the equity portion - and that would be the case regardless of how the equity portion is allocated. The return will be whatever it will be, and I will need to accept that. My focus is on limiting the downside.
Last edited by Call_Me_Op on Thu Nov 23, 2017 9:43 am, edited 1 time in total.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein

User avatar
nedsaid
Posts: 9524
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by nedsaid » Thu Nov 23, 2017 9:43 am

My reaction to Larry's article is that there really is not a science of investing. I believe you can use certain scientific and statistical techniques to learn what works and what doesn't work. You also get keen insights as to why certain things work. While not a subjective "soft" science, the "science of investing" isn't really a hard science either. It just is not "plug and chug" physics. Why do I make these statements? The wild card of human irrationality at precisely the wrong moment, in other words, human nature and behavior screws this up to a degree. The markets will do precisely what the experts do not expect, particularly during a crisis.

The scientific method involves coming up with a hypothesis and testing it with experimentation and the resulting data. If you have a scientific law, you ought to be able to recreate the results over and over and over again. You can do tests in the present. The "science of investing" is no such thing, it works with past data but you cannot project the results into the future. While a scientific experiment can have the results replicated over and over and over again, with investing we are told past performance is no guarantee of future results.

Pretty much, what the "science of investing" can tell us is what works most of the time. But as we found out during the 2008-2009 financial crisis, most everything fell and fell hard. Even "safer" investments like investment grade corporates and TIPS fell hard. Only nominal treasuries and certain US Government Agency bonds did well in that environment. Even GE, with its AAA credit rating at the time could not roll over its commercial paper. Warren Buffett bailed out GE. So managing risk with factors during the 2000-2002 bear market worked well but was a big fail in 2008-2009. It was the wild card of human behavior, fear and downright irrationality, struck at the worst possible time and blew all of this up. In a panic, investors sent the baby, the bathwater, and the wash tub all sailing out the window.

So while Larry's "science of investing" is a big improvement over what existed before, it is not perfect as all the math does is describe human behavior in markets. Pretty much this is putting the market on the psychiatrist's couch and putting numbers to it.
A fool and his money are good for business.

Da5id
Posts: 2035
Joined: Fri Feb 26, 2016 8:20 am

Re: Larry Swedroe: Managing Risk With Factors

Post by Da5id » Thu Nov 23, 2017 9:47 am

in_reality wrote:
Wed Nov 22, 2017 9:55 pm
I truly do wish factor investing was logically consistent, and didn't try co-opt the notion of diversification.
I think factors have something to them, though I haven't drunk the koolaid enough to do anything about them. But this usage bothers me too. I had an interaction with Larry before he left where he said he felt that a small cap value fund was more diversified than TSM. Because it held 2 factors (small, value) and TSM has one (beta). Small cap value may be better, but this appears to me to be a tendentious usage of the word "diversified". Factors should stand on their own without taking a positive word that means something and co-opting it.

I'll give a different contextual example of what I mean. Japanese and German cars are better (assume it) by some consistent measure. There is thus a positive "Japan" factor and a positive "German" factor for brands of cars in your fleet. Which of these fleets of cars is more "diversified" (looking at 15 top world manufacturers by production):

1) Fleet A
Toyota
VW
Nissan
Honda
Suzuki
Daimler
BMW

2) Fleet B
Toyota
VW
Hyundai
Honda
GM
Ford
Nissan
Honda
Suzuki
Fiat
Renault
PSA
SAIC
Daimler
BMW
Changan

Hmm. Obviously Fleet A is more "diversified", because it selects for the Japan factor and the German factor... Doesn't fly to me. Fleet A may have a higher concentration of better cars (I'd bet on it myself). But it is, to me, hijacking the word diversified. It feels like newspeak, concentration means diversification.

betablocker
Posts: 366
Joined: Mon Jan 11, 2016 1:26 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by betablocker » Thu Nov 23, 2017 10:04 am

Da5id wrote:
Thu Nov 23, 2017 9:47 am
in_reality wrote:
Wed Nov 22, 2017 9:55 pm
I truly do wish factor investing was logically consistent, and didn't try co-opt the notion of diversification.
I think factors have something to them, though I haven't drunk the koolaid enough to do anything about them. But this usage bothers me too. I had an interaction with Larry before he left where he said he felt that a small cap value fund was more diversified than TSM. Because it held 2 factors (small, value) and TSM has one (beta). Small cap value may be better, but this appears to me to be a tendentious usage of the word "diversified". Factors should stand on their own without taking a positive word that means something and co-opting it.

I'll give a different contextual example of what I mean. Japanese and German cars are better (assume it) by some consistent measure. There is thus a positive "Japan" factor and a positive "German" factor for brands of cars in your fleet. Which of these fleets of cars is more "diversified" (looking at 15 top world manufacturers by production):

1) Fleet A
Toyota
VW
Nissan
Honda
Suzuki
Daimler
BMW

2) Fleet B
Toyota
VW
Hyundai
Honda
GM
Ford
Nissan
Honda
Suzuki
Fiat
Renault
PSA
SAIC
Daimler
BMW
Changan

Hmm. Obviously Fleet A is more "diversified", because it selects for the Japan factor and the German factor... Doesn't fly to me. Fleet A may have a higher concentration of better cars (I'd bet on it myself). But it is, to me, hijacking the word diversified. It feels like newspeak, concentration means diversification.
To use your analogy assume there is are positive Japan and Germany factors but US manufacturers represent 95% of the market. If you held all cars by market size you’d effectively exposed to almost all US. The US is beta.

betablocker
Posts: 366
Joined: Mon Jan 11, 2016 1:26 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by betablocker » Thu Nov 23, 2017 10:04 am

Da5id wrote:
Thu Nov 23, 2017 9:47 am
in_reality wrote:
Wed Nov 22, 2017 9:55 pm
I truly do wish factor investing was logically consistent, and didn't try co-opt the notion of diversification.
I think factors have something to them, though I haven't drunk the koolaid enough to do anything about them. But this usage bothers me too. I had an interaction with Larry before he left where he said he felt that a small cap value fund was more diversified than TSM. Because it held 2 factors (small, value) and TSM has one (beta). Small cap value may be better, but this appears to me to be a tendentious usage of the word "diversified". Factors should stand on their own without taking a positive word that means something and co-opting it.

I'll give a different contextual example of what I mean. Japanese and German cars are better (assume it) by some consistent measure. There is thus a positive "Japan" factor and a positive "German" factor for brands of cars in your fleet. Which of these fleets of cars is more "diversified" (looking at 15 top world manufacturers by production):

1) Fleet A
Toyota
VW
Nissan
Honda
Suzuki
Daimler
BMW

2) Fleet B
Toyota
VW
Hyundai
Honda
GM
Ford
Nissan
Honda
Suzuki
Fiat
Renault
PSA
SAIC
Daimler
BMW
Changan

Hmm. Obviously Fleet A is more "diversified", because it selects for the Japan factor and the German factor... Doesn't fly to me. Fleet A may have a higher concentration of better cars (I'd bet on it myself). But it is, to me, hijacking the word diversified. It feels like newspeak, concentration means diversification.
To use your analogy assume there are positive Japan and Germany factors but US manufacturers represent 95% of the market. If you held all cars by market size you’d be exposed to almost all US. The US is beta. Japan and Germany are value and momentum.

CULater
Posts: 1037
Joined: Sun Nov 13, 2016 10:59 am

Re: Larry Swedroe: Managing Risk With Factors

Post by CULater » Thu Nov 23, 2017 10:14 am

I repeat my previous query: if your "factor black box" doesn't work for a couple of years, how do you know if it's broken or if it's temporarily out of favor; and whether to get out or buy more by rebalancing? I just don't think you know without drinking more of the koolaid. Sounds like a real challenge to Bogle philosophy to me.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Larry Swedroe: Managing Risk With Factors

Post by in_reality » Thu Nov 23, 2017 10:51 am

Da5id wrote:
Thu Nov 23, 2017 9:47 am
I had an interaction with Larry before he left where he said he felt that a small cap value fund was more diversified than TSM. Because it held 2 factors (small, value) and TSM has one (beta).
Me too, but it doesn't seem to make sense that TSM has beta exposure (which can be divided into high and low) but not size (which can be divided into large and small).
betablocker wrote:
Thu Nov 23, 2017 10:04 am
To use your analogy assume there are positive Japan and Germany factors but US manufacturers represent 95% of the market. If you held all cars by market size you’d be exposed to almost all US. The US is beta. Japan and Germany are value and momentum.
If beta is singular, how can we divide it into low and high beta stocks to form the BAB factor? In fact we should.

You have obviously split the market by style to derive your Japan factor, but the non-value holdings seem to have been lost in space.

Korea must be growth then, because if you have divided stock by style to form value, then growth stocks must exist too. I guess they have been obliterated by a mad dictator via nuclear weapons.

And let's assume China is the opposite of momentum because their economy gets bad press (and was down yesterday when every place else was flat).

Now you are saying that a portfolio of US, Japan and Germany is more diversified than the US, Japan, Germany, Korea and China.

Jettison Korea and China to space in your model if you wish I guess. Whatever.

Anyway, the claim of greater diversification is simply an artifact of the measurement system. [maybe I should "byproduct" instead of "artifact", would that make it easier for you to understand]

In fact, Japan in your model doesn't represent the earnings from Japanese automakers, it represents the additional profit Japanese car manufactures earn over Korean ones. And you are suggesting we are better diversified for ignoring Korea. Seriously whatever. Gamble on Japan if you want, but calling it diversification is absurd.

User avatar
GreatOdinsRaven
Posts: 524
Joined: Thu Apr 23, 2015 8:47 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by GreatOdinsRaven » Thu Nov 23, 2017 11:25 am

in_reality wrote:
Thu Nov 23, 2017 10:51 am
Da5id wrote:
Thu Nov 23, 2017 9:47 am
I had an interaction with Larry before he left where he said he felt that a small cap value fund was more diversified than TSM. Because it held 2 factors (small, value) and TSM has one (beta).
Me too, but it doesn't seem to make sense that TSM has beta exposure (which can be divided into high and low) but not size (which can be divided into large and small).
Wouldn’t a SCV fund (for example DFA’s) have exposure to three factors: small, value and beta? 0.99 loading on beta, right?

https://www.portfoliovisualizer.com/fac ... sion=false

Regarding TSM it has a beta of 1 by definition. If you want low beta then you are by definition tilting away from the higher beta stocks in a TSM portfolio and therefore it’s no longer a market-cap weighted TSM portfolio.
"The greatest enemies of the equity investor are expenses and emotions." -John C. Bogle, Little Book of Common Sense Investing. | | "Winter is coming." Lord Eddard Stark.

User avatar
packer16
Posts: 951
Joined: Sat Jan 04, 2014 2:28 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by packer16 » Thu Nov 23, 2017 11:31 am

One issue with factors for diversification is timing. We have observed on average factors provide a diversification benefit but it is time dependent. OTOH, bonds provide diversification always. So when you substitute factors for bonds or have a more bonds with a higher expected return there is a high probability that you will be disappointed if time horizon is short, less than 10 to 20 years. If your horizon is that long then IMO the benefits of diversification are small as you will not pulling any money out until the end & the benefit is mainly psychological (you do not like to see volatile returns).

Another issue is the fact the some factors are going to provide you a better risk /reward mix that the index. Is it not implicit in this statement that the market is not providing the best average risk/reward with market cap weighting. So, similar to active investing, you are betting the market is wrong on a macro basis. This is the hardest part for me to accept that you are on macro basis going to do better than market weights. I can buy this for some small niches but it is much harder to accept the market as a whole is on average mispricing factor stocks.

Packer
Buy cheap and something good might happen

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 11:47 am

In_reality,
I think your basic understanding of factors has a mistaken premise. You’ve mentioned that large should be uncorrelated with small, profitable uncorrelated with unprofitable, growth uncorrelated with value. That is 180 degrees off target. They are not only correlated, they are the same thing. They are measures of unique independent qualities separate from the market factor. It’s like weight: something can be heavy or light, but the same thing is being measured. Or better yet, elevation. You can have both positive values in the air and negative values under the sea, but the same thing is being measured.
The factors are long-short portfolios. For example value is top top 30% minus bottom 30% by book to market. That equation eliminates the market factor and isolates the return to the price factor. One can have a very heavy positive load on the price factor, a tilt to value. Or he can have a heavy negative load on the price factor, a tilt to growth. Or he can be neutral on the spectrum with TSM. But to say value/growth, small/big, low investment/high investment are uncorrelated pairs is incorrect. It’s the opposite. They are measures of the same thing: a unique factor explanatory of returns independent of the market factor and the other factors. One can have positive, zero, and negative exposure to the various factors.

Dave

stlutz
Posts: 4469
Joined: Fri Jan 02, 2009 1:08 am

Re: Larry Swedroe: Managing Risk With Factors

Post by stlutz » Thu Nov 23, 2017 11:51 am

To bring this discussion back on track: The definition of a risk factor is that it measures *concentration* of risk, not *diversification* of it.

If you analyze your portfolio against a bunch of factors (market, valuation, term, credit, sector etc.) and end up with a bunch of numbers close to zero, you have a highly diversified portfolio in terms of where your risk is coming from. If you have a bunch of highly positive or negative numbers, your risk is concentrated--the more big numbers you see the more concentrated you are.

If my portfolio is 100% TSM, I'll have a market factor of 1.0. That simply indicates that my risks are concentrated entirely in the stock market as opposed to in a variety of assets (bonds, rental homes, commodities etc.) .

If my portfolio is instead 100% Vanguard Smallcap Value, I now have 3 big factor loadings--market at .99, size at .51 and value at .6. Did my portfolio become more diversified? No. I still have the same market exposure, but now I've added two additional big numbers. My portfolio has become *more* concentrated, not less.

Suppose I instead did something like 1/3 TSM, 1/3 Total Bond, and 1/3 commodities (PCRIX). Now I'll get my market exposure down to .45. I've added term and credit at .21 and .40. So, I've reduced my huge exposure that I had before, but I've added some newer concentrations that aren't as big. So I've cut out the huge market risk, but added some term/credit risks. On balance I've become more diversified, but there is more I could do if I was seeking maximum diversification.

Is this portfolio *better* than my 100% TSM portfolio? That's debatable. Constructing a good portfolio involves trade offs. Diversification is a good thing. It's not the only thing. Same with volatility, one's return forecasts etc.

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 11:52 am

Call me op,
Have you ever used Monte Carlo simulation? I’ve found it very enlightening. The expected draw down is the output. By necessity one has to have inputs with some estimation of expected returns. How does one rationally choose a basic stock bond split much less tilts without some estimate of returns required /desired by the investor to meet his goals?

Dave

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 12:03 pm

Nedsaid,
Totally agree that investing is not physics. Human behavior really jacks finance up. But that is a large part of what makes investing so fascinating.
That being said, it’s not all hopeless. People have looked at data, made hypotheses, and evaluated their hypotheses with out of sample data. For example I believe Fama French original work was US 1963-1992 or something like that. The size and value factors have subsequently been evaluated in different time periods and different geographic markets. So, while it may not be physics, it’s certainly a far cry from astrology :-)

Dave

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 12:11 pm

CULater,
While I don’t see factor investing as a black box, you make a very good point. When a factor performs poorly for an extended period of time, is it just a bad streak or has it been permanently eradicated? No way to know until after the fact! We can only invest looking forward. That’s why staying committed to a factor requires a strong forward looking risk based or behavioral based explanation that the investor believes in. I think when discussing factors on this forum, people put too much weight on historical performance and don’t put enough weight on the intuitive rationale for sticking to the factors in the future. For example we always hear about past performance of value, but how often do we mention that the cost of capital for a distressed firm is likely higher than for a big growth one?

Dave

User avatar
nedsaid
Posts: 9524
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by nedsaid » Thu Nov 23, 2017 12:17 pm

Random Walker wrote:
Thu Nov 23, 2017 12:03 pm
Nedsaid,
Totally agree that investing is not physics. Human behavior really jacks finance up. But that is a large part of what makes investing so fascinating.
That being said, it’s not all hopeless. People have looked at data, made hypotheses, and evaluated their hypotheses with out of sample data. For example I believe Fama French original work was US 1963-1992 or something like that. The size and value factors have subsequently been evaluated in different time periods and different geographic markets. So, while it may not be physics, it’s certainly a far cry from astrology :-)

Dave
That is my take as well, not physics but a far cry from astrology. As long as one realizes that the "science of investing" isn't perfect and that no investment strategy works all the time, one is on solid ground. This is why I am not in the land of "always" or "never." Successful investing is about shifting the odds in your favor as much as possible but obviously there are no guarantees.
A fool and his money are good for business.

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 12:19 pm

Packer,
I believe the market is super efficient. In fact I probably believe it’s more efficient than it actually is :-). The market prices risk. Small and value have been shown to be separate, unique, independent risks from the market factor. For a single stock, the market weights these unique different risks and comes up with a single number, the stock’s price. So saying that there are small and value premia is not inconsistent with the market’s efficient pricing of risk(s).

Dave

User avatar
Taylor Larimore
Advisory Board
Posts: 26971
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Improving the "Efficient Portfolio" ?

Post by Taylor Larimore » Thu Nov 23, 2017 2:20 pm

Random Walker wrote:
Thu Nov 23, 2017 12:19 pm
Packer,
I believe the market is super efficient. In fact I probably believe it’s more efficient than it actually is :-). The market prices risk. Small and value have been shown to be separate, unique, independent risks from the market factor. For a single stock, the market weights these unique different risks and comes up with a single number, the stock’s price. So saying that there are small and value premia is not inconsistent with the market’s efficient pricing of risk(s).

Dave
Dave:

Inasmuch as you and Nobel Laureate, Bill Sharpe, agree that the efficient (highest return with least risk) portfolio has to be the market portfolio. It seems to me that adding (overweighting) other stocks or funds MUST make a portfolio less efficient.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

Random Walker
Posts: 2666
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by Random Walker » Thu Nov 23, 2017 3:12 pm

Taylor,
I don’t know much about Sharpe. I do understand the mathematics of active management story though. I think the behavior of individual portfolios may be where we differ. I’m a huge fan of modern portfolio theory: how different less than perfectly correlated sources of return improve portfolio behavior.

Dave

User avatar
packer16
Posts: 951
Joined: Sat Jan 04, 2014 2:28 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by packer16 » Thu Nov 23, 2017 10:29 pm

I have the same understanding as Taylor. If the market efficiently prices all risk including the factor risks, then would not the market weight represent the markets best estimate of the optimal risk/reward ratio? If not, then why not? IMO by factor investing at different weights than those set by the market you are stating the weights you are choosing is more efficient than market weights based upon many market participants. I can buy that if you focus on a few securities you can find a few mispricings or non-optimal risk/return situations but investing as though there is a cross sectional mispricing I think is a big assumption. The excess return you may be seeing is due to timing risk associated with factors & I see no way to remove or diversify the timing risk.

Packer
Buy cheap and something good might happen

User avatar
nedsaid
Posts: 9524
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by nedsaid » Fri Nov 24, 2017 12:21 am

packer16 wrote:
Thu Nov 23, 2017 10:29 pm
I have the same understanding as Taylor. If the market efficiently prices all risk including the factor risks, then would not the market weight represent the markets best estimate of the optimal risk/reward ratio? If not, then why not?

Nedsaid: Except that markets do exhibit periods of irrationality, where markets are too optimistic or too pessimistic. Euphoria and panic. Does anyone really believe that the high flying internet and high tech stocks were efficiently priced during the late 1990's mania? Companies with no earnings and in a few cases, no sales were priced pretty much to infinity. This is efficient pricing? Really?

To me, this is the whole point of Value investing, taking advantage of mispricing in the markets. Just as entire markets can experience bouts of irrationality, pricing of individual securities might not be 100% efficient. Isn't that the point of Benjamin Graham and his famous disciple Warren Buffett calculating intrinsic value for companies? The efficient market hypothesis assumes that human beings are rational and yet market history shows time and time again that this just ain't so.

I will say that the collective judgment of the participants in financial markets is mostly very good. Markets are for the most part pretty darned efficient. But again, it goes back to what I argued above in this thread. "Investing science" is really about what seems to work most of the time except for those times when those darned humans don't behave as expected. This is why I don't live in an "always" or "never" world. No investing strategy works all of the time, even the very best strategies work most of the time.

The thing is, Value as a long term strategy is a proven and effective strategy. A lot of great data to support it. Most of the famous investment managers that I can think of were Value managers or at least Value oriented.

The problem is that the market has favored the Large Growth stocks since the 2008-2009 financial crisis. Part of it is flight to quality and part of it has to do with a slower growth economy and very low interest rates which have contributed to the market putting a greater premium on consistent growth, wherever the market can find it. Since you can't get a decent return from low risk bonds, investors are looking to quality growth stocks for that return even if they have to take more risk to get it.

I have read many of Packer16's posts. He was an eloquent proponent of the Benjamin Graham style of investing and one of the usual suspects in the Value threads. It appears he has thrown in the towel on Value investing and wants to just index his money. Value just hasn't performed since the financial crisis and now its most outspoken advocate here on the forum has thrown in the towel. Pretty much, the last optimist has become pessimistic and this makes me suspect that good times are ahead for Value investors. Next, I expect Larry Swedroe to tearfully confess that he was all wrong about factors and has 100% of his monies in the Taylor Larimore 3 fund portfolio. I volunteer to send Packer and Larry their crying towels. Somehow, I don't think Larry will fold. I don't know, maybe we will learn that Buffett and Munger just threw darts at a stock page and that accounted for Berkshire-Hathaway's amazing success. Don't think that will happen either.


IMO by factor investing at different weights than those set by the market you are stating the weights you are choosing is more efficient than market weights based upon many market participants. I can buy that if you focus on a few securities you can find a few mispricings or non-optimal risk/return situations but investing as though there is a cross sectional mispricing I think is a big assumption. The excess return you may be seeing is due to timing risk associated with factors & I see no way to remove or diversify the timing risk.

Nedsaid: I will say that markets go through phases. We are in a Large Growth market right now so Value doesn't look so good right now. This might be the timing risk that Packer is talking about. The thing is, the academics find Value premiums all over the world. It isn't a phenomenon that exists only here in the United States. This makes me suspect that human nature and human behavior is involved in this. I don't know, maybe the problem is that humans have been too rational recently. We haven't seen irrationality since 2008-2009. We will see. My suspicion is that Value will be back, and with a vengeance. It will happen just after Packer puts his every dime into Large Growth stocks. We will see.


Packer
A fool and his money are good for business.

Call_Me_Op
Posts: 6815
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry Swedroe: Managing Risk With Factors

Post by Call_Me_Op » Fri Nov 24, 2017 7:11 am

Random Walker wrote:
Thu Nov 23, 2017 11:52 am
Call me op,
Have you ever used Monte Carlo simulation? I’ve found it very enlightening. The expected draw down is the output. By necessity one has to have inputs with some estimation of expected returns. How does one rationally choose a basic stock bond split much less tilts without some estimate of returns required /desired by the investor to meet his goals?

Dave
Yes, I am an engineer, so I have used Monte Carlo for many applications.

My point is that expected return is not that important for me. I want to obtain the best return I can while limiting the downside. The downside protection take precedence in my case. I know that this is not the case for many.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein

User avatar
packer16
Posts: 951
Joined: Sat Jan 04, 2014 2:28 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by packer16 » Fri Nov 24, 2017 7:44 am

To be clear I have not left value investing. I still believe and a vast majority of my portfolio (90%+) is in invested in value stocks. However, the value factor is not value investing & conflating the two leads all types of confusion. Value investing can work due to a small number of misprcings. However, the premise of the value factor is there is widespread cross-sectional either mispricing or improved risk/return relationship across value factor stocks. It this premise (specifically the widespread aspect) that I have an issue with. Value tilters are not practicing value investing in the traditional Graham & Dodd sense. They are using a trading strategy that has proven effective with past data & extrapolating it to the future. This is very different from bottoms valuation of securities.

Packer
Buy cheap and something good might happen

User avatar
Robert T
Posts: 2518
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Larry Swedroe: Managing Risk With Factors

Post by Robert T » Fri Nov 24, 2017 7:59 am

.
On the market-portfolio being mean-variance efficient (i.e. "highest return with least risk") – is true if you believe the Capital Asset Pricing Model (Sharpe). It is not true if you believe in multi-factor asset pricing models (such as the Fama-French three factor model). In a multi-factor world other things matter beyond just sensitivity to market volatility, such as sensitivity to recession risk. The CAPM (sensitivity to market volatility) performs poorly at explaining the variation of returns of value and small cap stocks, relative to multi-factor models.

Johan Cochrane has an excellent earlier article (1999) on this point. http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf
  • “Now, investors care about three attributes of their portfolios: 1) They want higher average returns; 2) they want lower standard deviations or overall risk; and 3) they are willing to accept a portfolio with a little lower mean return or a little higher standard deviation of return if the portfolio does not do [as] poorly in recessions.”

The point above I underlined implies a lower Sharpe ratio (lower mean return or higher SD) but less relative poor performance in recessions. This is what we observe if we compare the returns of Vanguard Total Stock Market (VTSMX) with iShares S&P 600 Value (IJS), since the inception of IJS in August 2000:

https://www.portfoliovisualizer.com/bac ... ion2_2=100
  • Sharpe / Max drawdown
    VTSMX = 0.36 / -50.9
    IJS = 0.55 / -54.1
Obviously no guarantees – but Cochrane's article above is worth a read for those interested.

Robert
.

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Larry Swedroe: Managing Risk With Factors

Post by in_reality » Fri Nov 24, 2017 10:44 am

Robert T wrote:
Fri Nov 24, 2017 7:59 am
.
On the market-portfolio being mean-variance efficient (i.e. "highest return with least risk") – is true if you believe the Capital Asset Pricing Model (Sharpe). It is not true if you believe in multi-factor asset pricing models (such as the Fama-French three factor model). In a multi-factor world other things matter beyond just sensitivity to market volatility, such as sensitivity to recession risk. The CAPM (sensitivity to market volatility) performs poorly at explaining the variation of returns of value and small cap stocks, relative to multi-factor models.

Johan Cochrane has an excellent earlier article (1999) on this point. http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf
  • “Now, investors care about three attributes of their portfolios: 1) They want higher average returns; 2) they want lower standard deviations or overall risk; and 3) they are willing to accept a portfolio with a little lower mean return or a little higher standard deviation of return if the portfolio does not do [as] poorly in recessions.”

The point above I underlined implies a lower Sharpe ratio (lower mean return or higher SD) but less relative poor performance in recessions. This is what we observe if we compare the returns of Vanguard Total Stock Market (VTSMX) with iShares S&P 600 Value (IJS), since the inception of IJS in August 2000:

https://www.portfoliovisualizer.com/bac ... ion2_2=100
  • Sharpe / Max drawdown
    VTSMX = 0.36 / -50.9
    IJS = 0.55 / -54.1
Obviously no guarantees – but Cochrane's article above is worth a read for those interested.

Robert
.
It's a very good read. Cochrane suggests those not holding value stocks are paying a premium to do so, and suggests those who are not sensitive to recessions risk (and can sell insurance to those willing to buy) "should buy extra amounts of recession sensitive stocks, value stocks, high yield bonds, etc., if these strategies carry a credible high average return. This action works just like selling insurance, in return for the premium".

That's consistent with what I've read people increase value exposure as they get older, wealthier, have less macroeconomic risk and heard it explained as having less need to hedge employment risks. **

The idea of selling insurance makes sense to me in context of the Larry portfolio which I think does have a risk of underperforming if the small value doesn't do quite well. If you have a fairly big portfolio though, the security of the bonds and downside protection may be worth it as the potential lack of growth is a chance you can take, especially small value could come through.

I agree with Cochrane that "The extra factors and time-varying returns*** would not be there (and will quickly disappear in the future) if lots of people were willing and able to take them."

And I also should probably consider that:
Do not forget, the average investor holds the market. If you'’re pretty much average, all this thought will lead you right back to holding the market index. To rationalize anything but the market portfolio, you have to be different from the average investor in some identifiable way. The average investor sees some risk in value stocks that counteracts their attractive average returns. Maybe you should too!
Perhaps Bogleheads are wealthier with less macroeconomic exposure than average and can afford to sell insurance.

In any case, I particularly like Cochrane's discussion of risks and his statement that "Unfortunately, the arguments that a factor will
persist are all inconsistent with aggressive portfolio advice" and that "there are several important qualifications that should
temper one’s enthusiasm and that shade portfolio advice back to the traditional view" and away from constructing portfolios based on multiple factors.

So I think Cochrane's article is saying that after consideration of the risks behind premiums in multiple factor portfolios, that the default should be something like a typical Boglehead three fund porfolio such as Taylor Larimore proposes.


** for instance. http://www.mysequoiawealth.com/the-char ... investors/
*** meaning the profitability of market timing

User avatar
nedsaid
Posts: 9524
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by nedsaid » Fri Nov 24, 2017 11:34 am

packer16 wrote:
Fri Nov 24, 2017 7:44 am
To be clear I have not left value investing. I still believe and a vast majority of my portfolio (90%+) is in invested in value stocks. However, the value factor is not value investing & conflating the two leads all types of confusion. Value investing can work due to a small number of misprcings. However, the premise of the value factor is there is widespread cross-sectional either mispricing or improved risk/return relationship across value factor stocks. It this premise (specifically the widespread aspect) that I have an issue with. Value tilters are not practicing value investing in the traditional Graham & Dodd sense. They are using a trading strategy that has proven effective with past data & extrapolating it to the future. This is very different from bottoms valuation of securities.

Packer
Darn it, you didn't throw in the towel after all so we might have to wait for Value to come back.

We have discussed the difference between the Benjamin Graham approach and the Academic approach and yes, they are far different. My understanding of Value is closer to Benjamin Graham than the academics. What DFA is now doing is screening for Value and setting momentum to neutral. My understanding is that they also screen for Profitability/Quality which makes their approach now more Buffett like.
A fool and his money are good for business.

User avatar
abuss368
Posts: 12408
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Larry Swedroe: Managing Risk With Factors

Post by abuss368 » Fri Nov 24, 2017 1:37 pm

Rick Ferri wrote:
Tue Nov 21, 2017 4:34 pm
GreatOdinsRaven wrote:
Mon Nov 20, 2017 10:11 pm
But, I'm still a Boglehead. As Rick Ferri has said many times, most of us are Bogleheads. That's our philosophy. We like passive and index funds. We like low cost and low turnover. We often (?usually?) like static strategic allocations and buy and hold investing. How we implement our portfolio allocation is our individual strategy and with few exceptions none of us have the exact same strategy or portfolio. And, that's ok.
I'm happy this concept is being discussed. I call it “The Anatomy of Successful Index Investing." The process works when three important elements are present:

1) Philosophy: be a Boglehead, i.e. believe in low fees, low turnover, passive concepts meaning don’t try to beat the market. This is universal – we’re all Bogleheads’.
2) Strategy: how you apply the philosophy to your unique situation. It’s about creating and implement a strategy the works well for you (there’s no perfect strategy). I’ve never found two Bogleheads who have the exact same portfolio. Everyone needs to figure this out for themselves.
3) Discipline: stay the course, don’t be swayed to chase whatever the flavor of the day is (which is often discussed on this forum). Discipline requires a set a rules to follow called an Investment Policy, and a way to stay focused on the philosophy. This forum is a good way to stay focused.

Rick Ferri
Hi Rick -

It is great to see you on the forum again. Hopefully you are heading back in this direction with your awesome investment advice!
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
grap0013
Posts: 1880
Joined: Thu Mar 18, 2010 1:24 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by grap0013 » Fri Nov 24, 2017 2:00 pm

vencat wrote:
Wed Nov 22, 2017 1:51 pm
Dave,
Can you give me a comparative alt portfolio (including MOM, Reinsurance etc.) to a 60:40 three fund one. I really want to follow this prospectively.
The alt portfolio has good 'science' but I'm skeptical of costs, implementation and performance in the next bear market and and I'm not even going in to tax considerations.
Unfortunately with investing, I believe, one has to embrace and accept 'fat tails' with lots of patience, discipline and a good dose of bonds.
I got you vencat: https://www.portfoliovisualizer.com/bac ... tion9_2=30

Real funds real fees. Tilt equity to the gills. Swap regular bonds with treasuries. Take 10% alternatives out from each stocks and bonds. The tilted/alternative portfolio has less severe and shorter duration drawdowns. Higher annualized return to boot.
Last edited by grap0013 on Fri Nov 24, 2017 2:12 pm, edited 1 time in total.
There are no guarantees, only probabilities.

vencat
Posts: 207
Joined: Thu Sep 10, 2009 6:30 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by vencat » Fri Nov 24, 2017 2:11 pm

Thanks for the info.

User avatar
TD2626
Posts: 609
Joined: Thu Mar 16, 2017 3:40 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by TD2626 » Fri Nov 24, 2017 6:40 pm

grap0013 wrote:
Fri Nov 24, 2017 2:00 pm
vencat wrote:
Wed Nov 22, 2017 1:51 pm
Dave,
Can you give me a comparative alt portfolio (including MOM, Reinsurance etc.) to a 60:40 three fund one. I really want to follow this prospectively.
The alt portfolio has good 'science' but I'm skeptical of costs, implementation and performance in the next bear market and and I'm not even going in to tax considerations.
Unfortunately with investing, I believe, one has to embrace and accept 'fat tails' with lots of patience, discipline and a good dose of bonds.
I got you vencat: https://www.portfoliovisualizer.com/bac ... tion9_2=30

Real funds real fees. Tilt equity to the gills. Swap regular bonds with treasuries. Take 10% alternatives out from each stocks and bonds. The tilted/alternative portfolio has less severe and shorter duration drawdowns. Higher annualized return to boot.
Be careful backtesting with short time periods. While that's an interesting backtest, 4 years is nowhere near enough to evaluate something like this in my opinion. I usually try to perform multi-decade backtests. I feel that hasn't been around long enough to backtest that far back in the past is usually too new or too risky to be reasonable. There are limitations to backtesting, and history is no guarantee - even a long history, but especially a short history.

antiqueman
Posts: 440
Joined: Thu Mar 12, 2009 5:22 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by antiqueman » Sat Nov 25, 2017 6:55 pm

I don't understand the constant attack on those who wish to tilt to something other than the three fund portfolio. Its as if the three fund portfolio is the scared cow . There are post after post after post where people how bring up factors and are attacked. No wonder Larry left.

User avatar
patrick013
Posts: 2198
Joined: Mon Jul 13, 2015 7:49 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by patrick013 » Sat Nov 25, 2017 9:02 pm

Image

Having many stocks and bonds is certainly diversification. So, I add
some concern to reducing risk metrics like beta or standard deviation.
In this case beta with the reduced expectation of future returns. It's
not a Monte of course.
age in bonds, buy-and-hold, 10 year business cycle

User avatar
Robert T
Posts: 2518
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Larry Swedroe: Managing Risk With Factors

Post by Robert T » Sat Nov 25, 2017 9:05 pm

.
FWIW - some analysis on a hypothetical question:

Would it be ‘better’ to shift all equities in a portfolio to small value and correspondingly increase the bond allocation? The assumption is this would give similar/same expected return and lower volatility and tail risk.

The overall answer on ‘better or not” seems to depend on at least three things: (i) current equity holdings, (ii) ‘tracking error’ regret, and (iii) tolerable loss.

Here’s a comparison of two portfolios:
  • P1 = 75:25 stock/5 yr T-notes. Stock = 33% FF Large Cap Momentum: 33% Dimensional US Large Value: 17% Dimensional US Small: 17% Dimensional US Small Value. i.e. Dimensional US “Balanced” Equity with FF Large Cap Momentum in place of S&P500.

    P2 = 60:40 Dimensional US Small Value:5yr T-Notes.
A per the analysis below – the portfolios had similar annualized return over the 1928-2016 period. P1 (the 75:25 stock:bond portfolio) had lower standard deviation – so in this case shifting to a 60:40 SV:bond portfolio did not decrease portfolio volatility but increased it slightly (as reflected by SD). The portfolios had similar downsides in 1929-32, 1937, and 1973-74, with a larger difference in 2007-2008 where the 60:40 SV:bond portfolio provided 4% more downside protection (-19.4 vs. -23.5). During 1985-1999 when the S&P500 significantly outperformed, and tilting away from the market lagged, P1 had a 3.3% higher annualized return than P2 likely leading to less ‘tracking error’ regret.

To summarize – shifting to 100% SV for equities and increasing the bond allocation from 25% to 40%, (i) provided similar returns over the 1928-2016 period, but increased volatility (slightly higher SD and Sharpe ratio, lower mean-variance efficiency), (ii) provided similar downside protection in 1929-32, 1937, 1973-74, and 4% more protection in 2007-08, (iii) likely increased ‘tracking-error’ regret in periods such as 1985-1999.

Obviously no guarantees.

1928-2016: Annualized return (%)/SD
  • P1 = 11.4 / 18.2
    P2 = 11.3 / 19.1
P1 / P2 (cumulative declines %, nominal)
  • 1929-32 = -59.3 / -60.8
    1937 = -30.3 / -29.7
    1973-74 = -24.8 / -23.4
    2007-2008 = -23.5 / -19.4
1985-1999 (15 years): Annualized return
  • P1 = 16.6
    P2 = 13.3
    S&P500 = 18.9
.

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Larry Swedroe: Managing Risk With Factors

Post by in_reality » Sat Nov 25, 2017 9:18 pm

antiqueman wrote:
Sat Nov 25, 2017 6:55 pm
I don't understand the constant attack on those who wish to tilt to something other than the three fund portfolio. Its as if the three fund portfolio is the scared cow . There are post after post after post where people how bring up factors and are attacked. No wonder Larry left.
I personally apologize to everyone in the thread, especially Dave and Robert T, if my questioning in the thread is not productive.

RA looked at factor diversification globally and found:
Consistent with major asset classes, the cross-regional correlations within individual factors tend to rise during economic or market turbulence. Volatilities of the global equity factors rise during these periods and, portfolio diversification tends to weaken exactly when most needed. That said, long-term investors have reason to celebrate: in contrast to the upward trend in correlations documented for major asset classes, we find a more stable correlation structure among regional factor portfolios before and after the start of the millennium. All in all, in the factor space, diversification is alive and well.
https://www.researchaffiliates.com/en_u ... ctors.html

edge
Posts: 3300
Joined: Mon Feb 19, 2007 7:44 pm
Location: Great Falls VA

Re: Larry Swedroe: Managing Risk With Factors

Post by edge » Sat Nov 25, 2017 9:57 pm

Managing risk with factors gives too much credence to our ability to predict asset class behavior.

User avatar
TD2626
Posts: 609
Joined: Thu Mar 16, 2017 3:40 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by TD2626 » Sat Nov 25, 2017 11:21 pm

in_reality wrote:
Sat Nov 25, 2017 9:18 pm
antiqueman wrote:
Sat Nov 25, 2017 6:55 pm
I don't understand the constant attack on those who wish to tilt to something other than the three fund portfolio. Its as if the three fund portfolio is the scared cow . There are post after post after post where people how bring up factors and are attacked. No wonder Larry left.
I personally apologize to everyone in the thread, especially Dave and Robert T, if my questioning in the thread is not productive.
I certainly apologize as well if anyone - ever - is offended by any of my posts.

I don't think anyone should feel attacked in any thread, ever. Rational adults should be able to have a reasonable difference of opinion and hold a rational discussion about ambiguous issues. There are few certain truths in investing, and I welcome learning from even unconventional viewpoints. Indeed, I often respect members that hold unconventional viewpoints (even viewpoints I disagree with) for having the tenacity to stick with their convictions and post about it - hearing from many different types of investors makes this board an interesting place, instead of an echo chamber. So long as everyone's views are stated clearly and factually, and are evidence-based, there is no reason that disagreements can't be calm and civil.

User avatar
TD2626
Posts: 609
Joined: Thu Mar 16, 2017 3:40 pm

Re: Larry Swedroe: Managing Risk With Factors

Post by TD2626 » Sat Nov 25, 2017 11:49 pm

I think the point about "diversification" comes down to different definitions of the term.

Think of it like the difference between market-cap weighting and equal-weighting in stock funds.

Cap weighted funds buy an equal percentage stake in each company, while equal weight funds put an equal number of dollars in each company. Which is more "diverse"? It depends on your perspective.

Obviously, the equal weight portfolio would have a substantial small cap tilt, while the cap weight portfolio would be topheavy in a small number of the largest stocks on the market. However, there are issues with implementation of equal weight funds (costs, turnover, etc) and to me, ultimately an equal % stake in each company seems more diverse.

*Note - I sort of feel that market cap weighting should be called equal weighting (e.g. the terms are misnomers). Why? Market cap weights equal weight their percentage stakes in companies -- these funds buy an equal percent stake in each company (at least I think this is the case in general).

Post Reply