Swedroe: Volatility As A Strategy

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Random Walker
Posts: 2068
Joined: Fri Feb 23, 2007 8:21 pm

Re: Swedroe: Volatility As A Strategy

Post by Random Walker » Tue Aug 08, 2017 4:55 pm

Jbolden1517,
I think I understand your point about diversification and dilution. Is it fair to say that you would consider diversification of equities as adding another investment with equity like expected returns but low, non, negative correlation to equities? And dilution as adding in safer bonds with much lower volatility and expected return? I think that's just semantics, but point well taken. The rubber meets the road at the expected return, expected volatility, and cost of the portfolio as a whole.
When adding these alternatives to a portfolio, one needs to ask themselves what they are trying to accomplish. And that necessarily is intertwined with the question whether to create the position from the equity side, bond side, or some of both. In this light, Jbolden's diversification / dilution concept really represents a spectrum of tradeoffs. If one takes from the equity side alone, he is keeping expected return pretty much constant but decreasing portfolio volatility. I think Jbolden would consider this diversification of equity like risks. If one takes from the bond side, he is increasing expected return and increasing portfolio volatility. I think Jbolden would consider this decreased dilution by bonds and increased diversification of equity like risks. In both cases portfolio efficiency and portfolio cost are increased. Of course if one takes from both bonds and stocks, the effect is a combination of the two above.

Dave

jbolden1517
Posts: 868
Joined: Sun Jul 09, 2017 2:53 pm

Re: Bonds and Fixed Income vs. Alternate Stocks

Post by jbolden1517 » Tue Aug 08, 2017 6:08 pm

Theoretical wrote:EM Bonds - You've got an investability problem here. Do you pick USD Emerging bonds, which don't suffer currency problems on your end but are more likely to default (historically having a lower credit rating) or do you pick local currency bonds, which may have better credit but subject you to currency fluctuations that can easily outstrip the bonds' expected return.
That's get your non correlation high. That's what you want in diversifiers. High returning, volatile and non-correlating.
Theoretical wrote: International small - This is true, but the problem is that most international small cap funds are really international midcap funds at best. To truly get local exposure, you've got to pay ERs out the nose (well above 1%) and risk active management. Note, even DFA's is not a true ISV fund for many of its largest holdings.
Agree on the problem. The smaller you go the better. Same problem in the USA for microcaps and illiquidity. There are few good funds that go small enough. But mid is better than large. And as far as I'm concerned this is getting easier all the time. FNDC which I'm in is mostly $1-5b. About the same as DFA. I certainly wish mutual funds in general were better on size. OTOH when I'm buying or selling in the under $100m range myself I'm not uncommonly making market on one side for hours. If I can set the bid or ask, we aren't talking a lot of liquidity. So I get the problem.

That being said though every step down reduces correlation. Trying for it is at least worthwhile.
Theoretical wrote: Natural resources, [Precious metals equity], Energy, REITs - all are stock sectors or industries. REITs are a bit different, with the tax structure, but they still add complexity and TSM has a 3-4% slug.
I agree they are in there. But by overweighting you reduce portfolio risk.
Theoretical wrote: Gold and Commodities - 0% real return, with the main returns being speculative/political/event. You get storage costs and horrible tax treatment for gold and implementation problems/crowding with commodities.
On commodities I believe what Larry recommends is you look for commodities with backwardation so there is a substantial expected return. Right now incidentally that applies to gold (though this is rare).
Theoretical wrote: Moreover, with the exception of Gold, all of these assets tend to gain correlation in times of economic panic, such as the depression or 2008. Gold's actually not that great for inflation either.
Understood. 2008 was a lessen in how anti-correlation doesn't always work. After years of holding gold stocks for just such a moment... OTOH I made money in the 2000s bear, just sailed through, didn't do anything interesting. For people in TSM those 2 years of being ground down were very painful. It may not always work, but it works a lot. Sure you aren't protected against a global plague but you are protected against a dollar crash.

Are you really asserting that low or negative correlations don't exist?

jbolden1517
Posts: 868
Joined: Sun Jul 09, 2017 2:53 pm

Re: Swedroe: Volatility As A Strategy

Post by jbolden1517 » Tue Aug 08, 2017 6:16 pm

Random Walker wrote:Jbolden1517,
I think I understand your point about diversification and dilution. Is it fair to say that you would consider diversification of equities as adding another investment with equity like expected returns but low, non, negative correlation to equities? And dilution as adding in safer bonds with much lower volatility and expected return?
Dilution is adding cash. Diversification is adding a not perfectly correlating asset. Cash correlates with everything and nothing. A short term bond is essentially 98% better cash and 2% an interest rate future. The 2% is a diversifier, the 98% is a diluter.
Random Walker wrote: I think that's just semantics, but point well taken.
It is more than just semantics. Two portfolios;
A 10% stocks, 90% cash
B 200% stocks, -100% cash
A and B have different risk. But they perfectly correlate and have the same risk adjusted return. The difference between A and B is just dilution.

C 5% stocks, 5% EM bonds, 80% cash. C has lower risk than A but and it has a higher risk adjusted return. The difference between A and C is not just dilution there is diversification.

Theoretical
Posts: 1073
Joined: Tue Aug 19, 2014 10:09 pm

Re: Swedroe: Volatility As A Strategy

Post by Theoretical » Tue Aug 08, 2017 8:18 pm

Actually, you're understating cash here. Specifically, T-Bills have shown a pretty strong correlation to inflation shocks. Tyler 9000 did a pretty good analysis of cash, bonds and stocks and found that t-bills largely tracked the inflationary chaos in the 1970s in a way that neither bonds nor stocks did. Also the short and long ends of the yield curve look can look very different. Now this requires using a high yield savings account or rolling t-bills, but it's very different than a 0% checking account.

Taylor's preferred bond fund has a mix of maturities that average to intermediate but include both short and long, and it includes credit, treasury and agency debt.

A 3 funder gets:

US beta
International beta (with currency fluctuations that even out)
Term
Credit

In addition, he gets neutral (using 50/50) or negative (using 30/30) size exposure, which tampers down the volatility.

The other problem with the MPT non-correlation exercise is that it forces a lot more attention to the folio components to trim or boost that slice. It adds behavioral risk.

The problem with many of these middle-correlation assets is that they tend to only benefit (behaviorally) in stagnant/sideways markets. In bad times, either long treasuries or t-bills go bananas and everything else crashes. In good times, you get performance drag (witness all the threads about getting rid of International).

jbolden1517
Posts: 868
Joined: Sun Jul 09, 2017 2:53 pm

Re: Swedroe: Volatility As A Strategy

Post by jbolden1517 » Tue Aug 08, 2017 10:11 pm

Theoretical wrote:Actually, you're understating cash here. Specifically, T-Bills have shown a pretty strong correlation to inflation shocks.
Mostly I just consider you just subtract them off either side of the equation.
So the long bond X years = estimate FR over X years + duration risk premium
stocks real return = stock nominal return = nominal earnings growth + dividends - FR = real earnings growth + dividends
etc...

Image
Theoretical wrote: Tyler 9000 did a pretty good analysis of cash, bonds and stocks and found that t-bills largely tracked the inflationary chaos in the 1970s in a way that neither bonds nor stocks did.
I agree with that. Short term debt was popular for a reason when I was growing up. But mostly it just kept up with inflation. Investors can be too cautious after experiencing full force correlating risk head on. Bond investors were scarred by the experience.
Theoretical wrote: Also the short and long ends of the yield curve look can look very different. Now this requires using a high yield savings account or rolling t-bills, but it's very different than a 0% checking account.
Understood and agree.
Theoretical wrote: Taylor's preferred bond fund has a mix of maturities that average to intermediate but include both short and long, and it includes credit, treasury and agency debt.
I've gotten more positive towards TBM since I've been here. I used to be very much a go long or go short but there is nothing of value in the middle. Seems like for a long term holding, 60/40... TBM ain't much more risky than STB and the yield bonus helps quite a bit. So while I still might mix it up a bit more, let's say no major arguments there. Still think long zeros like Vanguard's EDV ETF might be very good non-correlating assets. Not so much opposite a large bond position but for the people who don't have one.
Theoretical wrote:A 3 funder gets:
US beta
International beta (with currency fluctuations that even out)
Term
Credit
The don't get much credit risk at all. Pretty close to none. On term they take on some, agree, but not much. Which is fine. Term risk can be scary about 30 basis points which can take a long time for compound to overwhelm bad timing. As for international beta you know that USA large and International large correlate too much. USA small, International small would get you two different betas. Again it is muted and suboptimal.

That pretty much is my problem Lots of USA beta and stuff cut with USA beta.
Theoretical wrote: In addition, he gets neutral (using 50/50) or negative (using 30/30) size exposure, which tampers down the volatility.
Agree with that. But at the cost of tampering down the returns quite a bit. We are testing a bunch of robo advisor portfolios at 70/30 vs. 3 fund 70/30 and its almost always well over100 basis points. You get to 50/50 vs 70/30 because 70/30 still has some nasty drawdown and yu just added another 80 basis points of lag. You'd be better off in a family of loaded high ER diversified mutual sold in a full service brokerage managed account at that point you've done so much damage to return.

Its one of the things that I've never quite gotten around here. People will freak about the risk of a portfolio under performing the SP500 by .5% or 1% and then hold 20% of their portfolio in cash as an "emergency fund".
Theoretical wrote: The other problem with the MPT non-correlation exercise is that it forces a lot more attention to the folio components to trim or boost that slice. It adds behavioral risk.
True in general. The robo-advisors all do that for people who don't want to. There are funds that use MPT but most have an active component.
Theoretical wrote: The problem with many of these middle-correlation assets is that they tend to only benefit (behaviorally) in stagnant/sideways markets. In bad times, either long treasuries or t-bills go bananas and everything else crashes.
I think you may be too scarred by 2008 when nothing worked. The whole world had to take the same beating regardless of what they invested in. That's the exception not the rule. There are crisis all the time internationally, USA investors don't notice. They don't notice because the low correlations work. They also don't notice when those markets skyrocket and they lag. Conversely 2000-2 was very comfortable for me I was up and in calm waters. I wasn't in tech so nothing much bad happened and lots of good happened.
Theoretical wrote: In good times, you get performance drag (witness all the threads about getting rid of International).
Part of what helps is that there aren't "good times" and "bad times". Every year 1/12 of the portfolio is getting just mauled. 1/6 are having a rough time. 1/2 is having a pretty time. 1/6 are booming and 1/12 is delivering crazy returns that have you worried it might be time to get out. You stop thinking of TSM bears as bad times and TSM bulls as good times.

With funds of course you are way more diversified (that's what you are paying for) so of course you correlate more. But for example consider a 70/30 Vanguard portfolio that just modifies 3 fund a bit:
14% / VIPSX -Vanguard Inflation Protected Securities
14% / BWX -SPDR Barclays Intrntl Treasury Bd
14% / VTSMX -Vanguard Total Stock Mkt Idx
14% / VIVAX -Vanguard Value Index
14% / VGSIX -Vanguard REIT Index
14% / VGTSX -Vanguard Total Intl Stock Index
14% / VGENX -Vanguard Energy

How much does that investor even notice most USA growth stock bear markets? Sure their TSM has a bad year but what does that do for the rest of the portfolio? How much lag do you see? You have to admit this portfolio is far safer than 70% TSM / 30% TBM and likely to perform better.

Sure I see all the threads about dumping international. I'm in tons of international (more than 1/2). I didn't have as good a 3 years as people concentrated in the USA. So far though the dollar selling off has been a nice unexpected boost. I like the valuations I'm seeing in that stuff. I also see the threads regarding concern about valuations in the USA market. Those pure TSM investors have every reason to be concerned. They have a lot of skin in the game for Facebook and Tesla. In the USA I was grabbing all sorts of secondary energy stocks and telcos last quarter when they were getting killed. These were hardcore bear market valuations, some of these stocks had double digit dividend yields, or close to it! Century Link, a global telco 13% dividend. Pipeline stocks at .3 book. For TSM this was a solid bull market. Now I have no idea if the pipeline oversupply problem clears up in 2019 or not. I could have a fantastic 2018 or not with that holding. But what I do know is whether USA pipeline clears up or doesn't sure doesn't correlate with whether Facebook and Tesla are having a good year. I don't have that fear at all about Facebook and Tesla. I don't have to care about what happens to TSM. Getting your correlations down really isn't hard and makes investing so much less scary. On top of substantially boosting serial return.

Hope that explains my POV on this.

jbolden1517
Posts: 868
Joined: Sun Jul 09, 2017 2:53 pm

Re: Swedroe: Volatility As A Strategy

Post by jbolden1517 » Wed Aug 09, 2017 7:16 am

Taylor Larimore wrote:
What would you be holding if you were 35 had an 8 year old. Had $85k for her education and had 10 years to make sure it got to $200? Bonds aren't going to do it.
jbolden1517"

First, I would recognize the fact that NO 85K investment will "make sure it got to $200 in ten years."

Second, I would not put 100% of the investment in stocks.

"What would you be holding?" I would probably put my 85K in a low-cost 529 tax-advantaged college plan containing low-cost Vanguard stock and bond index funds which increases its bond allocation as college years approach.
That increasing bond allocation increases his chances of missing the goal, it doesn't decrease the chances (unless of course the first 5 years are great in which case then the goal is slow or little growth). An investment plan is graded on how well it meets investment goals. What do you even mean by a good portfolio if you don't mean meeting peoples goals? The purpose of the money is to be able to do stuff with it. Its the stuff that's important, the investments are a means to an end.

I really am lost here in what you even mean by good investments.

grok87
Posts: 7572
Joined: Tue Feb 27, 2007 9:00 pm

Re: Swedroe: Volatility As A Strategy

Post by grok87 » Wed Aug 09, 2017 12:58 pm

jbolden1517 wrote:
Taylor Larimore wrote:
What would you be holding if you were 35 had an 8 year old. Had $85k for her education and had 10 years to make sure it got to $200? Bonds aren't going to do it.
jbolden1517"

First, I would recognize the fact that NO 85K investment will "make sure it got to $200 in ten years."

Second, I would not put 100% of the investment in stocks.

"What would you be holding?" I would probably put my 85K in a low-cost 529 tax-advantaged college plan containing low-cost Vanguard stock and bond index funds which increases its bond allocation as college years approach.
That increasing bond allocation increases his chances of missing the goal, it doesn't decrease the chances (unless of course the first 5 years are great in which case then the goal is slow or little growth). An investment plan is graded on how well it meets investment goals. What do you even mean by a good portfolio if you don't mean meeting peoples goals? The purpose of the money is to be able to do stuff with it. Its the stuff that's important, the investments are a means to an end.

I really am lost here in what you even mean by good investments.
I agree with Taylor.

In Unconventional Success, David Swensen provides a framework for investing for long-term goals, short term goals, and intermediate term goals:

1) Long term goals (10 years or greater):
50% Global Stocks
20% Real Estate
15% Treasuries
15% TIPs

2) Short term goals (2 years or less):
Cash/CDs

3) Intermediate Term goals (2-10 years): Interpolate between the two:
Example: Let's say you want to invest for a goal due in 6 years-for example college tuition payment. That is midway between Swensen's long-term=10 years and short-term=2 years. So you would hold this portfolio:

1/2 (50% Global Stocks, 20% Real Estate, 15% Treasuries, 15% TIPs) + 1/2 (100% Cash)=
25% Stocks
10% Real Estate
7.5% Treasuries
7.5% TIPs
50% Cash

And then as you move closer in time to the goal you would sell down the long term portfolio and move into cash until you are 100% in Cash or CDS when you are 2 years away.

cheers,
grok
"...people always live for ever when there is any annuity to be paid them"- Jane Austen

Post Reply