Harvey paper says re-balancing adds risk

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
VennData
Posts: 580
Joined: Mon Feb 26, 2007 4:52 pm

Harvey paper says re-balancing adds risk

Post by VennData »

I was led to this by another anti-rebalancing article on Marketwatch

http://www.marketwatch.com/story/the-hi ... 2014-12-09
While a routinely rebalanced portfolio such as a 60-40 equity-bond mix is commonly employed by many investors, most do not understand that the rebalancing strategy adds risk. Rebalancing is similar to starting with a buy and hold portfolio and adding a short straddle (selling both a call and a put option) on the relative value of the portfolio assets. The option-like payoff to rebalancing induces negative convexity by magnifying drawdowns when there are pronounced divergences in asset returns. The expected return from rebalancing is compensation for this extra risk. We show how a higher-frequency momentum overlay can reduce the risks induced by rebalancing by improving the timing of the rebalance. This smart rebalancing, which incorporates a momentum overlay, shows relatively stable portfolio weights and reduced drawdowns.
http://papers.ssrn.com/sol3/papers.cfm? ... id=2488552

While the caveat away the expected return from rebalancing, it seems to me the method of defining risk is the issue. So much academic finance uses options to understand the fundamentals of the question at hand, in this case it's the embedded volatility risk of the call/put.
User avatar
dodecahedron
Posts: 6607
Joined: Tue Nov 12, 2013 11:28 am

Re: Harvey paper says re-balancing adds risk

Post by dodecahedron »

I would prefer not to deal with rebalancing as I mentioned here, but the bottom line of this WSJ article about Harvey's paper is a bit suspect to me. It looks like they are promoting some kind of "smart rebalancing" involving momentum, a proprietary strategy promoted by Harvey's coauthors, who seem to work at a hedge fund. I am sure smart rebalancing doesn't come cheap!
final paragraph of MarketWatch article wrote:If you are a risk-averse shorter-term investor, he added, you need to be especially aware of rebalancing’s consequences. If you don’t want to increase your portfolio’s downside risk, you may want to avoid any rebalancing. If you nevertheless insist on doing so, Harvey said you should not mechanically rebalance at the end of every quarter or year, but instead time your rebalancings for periods when the markets’ trends appear to be reversing. He directs interested readers to his paper, which outlines one possible way of doing just that.


(Emphasis added by dodecahedron)
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Harvey paper says re-balancing adds risk

Post by in_reality »

I make no claim to knowing what is best but ...

Intuitively to me it has always seemed that rebalancing into stocks after a crash is definitely adding risk. That's it's likely to pay off eventually is another thing...
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

dodecahedron wrote:I would prefer not to deal with rebalancing as I mentioned here, but the bottom line of this WSJ article about Harvey's paper is a bit suspect to me. It looks like they are promoting some kind of "smart rebalancing" involving momentum, a proprietary strategy promoted by Harvey's coauthors, who seem to work at a hedge fund. I am sure smart rebalancing doesn't come cheap!
I am dubious of immediate rebalancing to target weights when stocks have fallen a lot. Stock market volatility rises when stocks fall, so your portfolio risk has already risen, even if your stocks are worth less than few months ago. My simple, free suggestion is to rebalance gradually. Do 10% of the projected rebalancing trade each month, instead of doing the entire trade immediately.
livesoft
Posts: 86075
Joined: Thu Mar 01, 2007 7:00 pm

Re: Harvey paper says re-balancing adds risk

Post by livesoft »

Many thanks for the link to the article and its link to the paper.
Wiki This signature message sponsored by sscritic: Learn to fish.
Khanmots
Posts: 1241
Joined: Sat Jun 11, 2011 2:27 pm

Re: Harvey paper says re-balancing adds risk

Post by Khanmots »

in_reality wrote:I make no claim to knowing what is best but ...

Intuitively to me it has always seemed that rebalancing into stocks after a crash is definitely adding risk. That's it's likely to pay off eventually is another thing...
Rebalancing is about managing risk.

Say you were at 50/50 equity/FI and that matched your need and ability to take risk.
The market crashes. You're now at 25/75. Your portfolio is now less risky than it was. You re-balance back to 50/50. Yes, you added risk... but only back to the point had previously decided was necessary and right.
Say the market now booms and you're at 75/25. You re-balance back to 50/50. You've removed risk... but again.. it's back to where you had previously determined you need to be.

I haven't read the paper (at work), but I suspect that it's only looking at one side of this.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

Here's a Dummies version of Harvey's work by Mark Hulbert. I've been stating this claim on the forum for years but now that it's been officially blessed with "research" maybe somebody will pay attention. Rebalancing when stocks go down is essentially a bet on mean reversion. Stocks are actually more risky when they're going down, not less risky. That's the whole problem with "rebalancing" theory. Risk (aka volatility) is not constant over time, so rebalancing to a fixed policy allocation doesn't make sense and never made sense as a "risk management" technique:
Whenever the market is in a longer-term up or down trend, rebalancing actually increases risk, particularly downside risk.

That’s because, when you rebalance, you take money away from the better-performing asset class and reinvest it in the poorer-performing one. If those two asset classes’ relative strength persists after the rebalancing, as they often do, you’ll end up worse off than if you had not rebalanced.

This new study, therefore, puts rebalancing in a whole new light. Rather than viewing it as entirely benign and unobjectionable, it needs to be seen as a risky bet that there will be a reversal in asset classes’ relative returns.

If you are willing to undertake more risk and have a very long-term horizon of at least 10 to 20 years, then rebalancing is an appropriate strategy. That’s because, over long periods of time, a periodically rebalanced portfolio is likely to earn enough to compensate you for its greater risk.
http://www.marketwatch.com/story/the-hi ... 2014-12-09
We don't know where we are, or where we're going -- but we're making good time.
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: Harvey paper says re-balancing adds risk

Post by rmelvey »

I rebalance religiously, but I understand the argument that it adds risk during an extreme environment. If any one of the assets goes to zero the whole portfolio approaches zero as you keep shoveling money into the asset that is declining. Think about a portfolio that is 50% Russia and 50% US stock market. If Russia's stock market goes to zero you will be throwing more money at it the whole way down and end up with a portfolio value close to zero.

We are conditioned to think that major indices won't go to zero, but they have in the past and certainly can in the future.
Last edited by rmelvey on Tue Dec 09, 2014 9:12 am, edited 1 time in total.
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Harvey paper says re-balancing adds risk

Post by nisiprius »

I don't know exactly how big a deal this actually amounts to. There have been a number of forum discussions about what I will call "rebalancing suckage." If you have a 60/40 portfolio and do not rebalance, and if bonds hold steady and stocks go to zero, you lose at most 60%. If you rebalance and stocks go to zero, theoretically you can lose 100% because the stocks keep sucking money out of your bonds as you keep throwing good money after bad.

On the one hand, this is a real effect and I really saw it during 2008-2009, when I had a lot of my portfolio in the Vanguard Balanced Index Fund, which continuously rebalances to 60/40. And, yes, I actually saw and was annoyed to see that if the S&P dropped by, say, 10%, my balanced index fund did not drop 60% of 10% = 6%, it dropped by noticeably more than that. Too lazy to run numbers, but maybe like 10% more--that is, when, based on mental math, I'd have expected my portfolio to be down 20%, it was really down 22%... that sort of thing.

On the other hand, when you do run the numbers it turns out that you need to be making almost apocalyptic assumptions about the severity of a stock decline for it to matter much.

If stocks don't actually go to zero, you make it all up on the way back--which is what happened in my Balanced Index. I didn't get any "rebalancing bonus" because the fund rebalances continuously. In order to get a rebalancing bonus you a) need to have mean reversion, and b) you need to have a long enough period of time between rebalancing for that mean reversion to take effect.

People (including me until I looked hard at it) make a mental error of imagining that rebalancing will magically include buying at the low. In reality, the rebalancing rules trigger at somewhat random points both on the way down and on the way up, and much of the hoped-for "buy low, sell high" effect cancels itself out--and the actual effect over any period of real history depends mostly on the luck of the timing of the rebalancing points and where they happen to land on the curve.
Last edited by nisiprius on Tue Dec 09, 2014 9:20 am, edited 2 times in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
letsgobobby
Posts: 12073
Joined: Fri Sep 18, 2009 1:10 am

Re: Harvey paper says re-balancing adds risk

Post by letsgobobby »

when stocks are moving up, rebalancing decreases risk (moving stocks to bonds). When stocks are moving down, rebalancing adds risk (moving bonds to stocks). As pointed out upthread, the goal is to maintain risk at a fairly constant level.
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: Harvey paper says re-balancing adds risk

Post by rmelvey »

letsgobobby wrote:when stocks are moving up, rebalancing decreases risk (moving stocks to bonds). When stocks are moving down, rebalancing adds risk (moving bonds to stocks). As pointed out upthread, the goal is to maintain risk at a fairly constant level.
You are not getting to the heart of the idea though because you are comparing apples and oranges (stocks and bonds). To understand why it adds risk you have to examine the dynamic of rebalancing between two assets with equal risk.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

rmelvey wrote:I rebalance religiously, but I understand the argument that it adds risk. If any one of the assets goes to zero the whole portfolio approaches zero as you keep shoveling money into the asset that is declining. Think about a portfolio that is 50% Russia and 50% US stock market. If Russia's stock market goes to zero you will be throwing more money at it the whole way down and end up with a portfolio value of zero.

We are conditioned to think that major indices won't go to zero, but they have in the past and certainly can in the future.
It's not just that the asset going down can keep going down -- it's the fact that stocks are actually more volatile during down markets than during up markets. Portfolio volatility is increasing. When you move money from less volatile bonds to more volatile stocks you're just exacerbating the problem. The only reason to do this is because you're betting on mean reversion. It's contrarian market timing pure and simple.
We don't know where we are, or where we're going -- but we're making good time.
letsgobobby
Posts: 12073
Joined: Fri Sep 18, 2009 1:10 am

Re: Harvey paper says re-balancing adds risk

Post by letsgobobby »

rmelvey wrote:
letsgobobby wrote:when stocks are moving up, rebalancing decreases risk (moving stocks to bonds). When stocks are moving down, rebalancing adds risk (moving bonds to stocks). As pointed out upthread, the goal is to maintain risk at a fairly constant level.
You are not getting to the heart of the idea though because you are comparing apples and oranges (stocks and bonds). To understand why it adds risk you have to examine the dynamic of rebalancing between two assets with equal risk.
I'm not going to read the paper because I don't have time, though I did read the article. For my benefit, what are the two assets with equal risk? ie, international and domestic stocks?

In the article, Hulbert misrepresents the cost because he states in 2013 rebalancing turned out to be a 'costly' move in that stocks continued to outperform in 2014. But that's not a cost any more than being 50% stocks instead of 60% stocks is a 'cost' - rather it's a direct reflection of receiving less reward for taking on less risk. And regardless, rebalancing from stocks to bonds at any time is not about increasing rewards, but about limiting downside risk.

On the downside, it is true - increasing stocks increases risk, which I agreed with.

However the article says rebalancing increases risks in both up and down markets. I'm hoping for an explanation of how risk is increased in an up market when stocks are sold and bonds are bought.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Harvey paper says re-balancing adds risk

Post by lack_ey »

The paper looks at 60/40 equity/bonds, with the premise based on momentum investing. Seems like some or most of the results use S&P 500 and 10-year treasuries to represent those components.

When looking at the period of 2000 to now and simulated market conditions, the proposed method of allocating a certain percentage (20%, 40%) of the portfolio to the momentum-based strategy improves performance. One could readily call data mining or overfitting data on this, as with most finance research. It's hard to know. There is also some mathematical analysis, but again the validity of that depends heavily on the assumptions made. A key idea is that during stock market downturns, rebalancing has you buying more stocks. The momentum investing does the opposite, so if you do both at the same time, the effect is somewhat canceled.

I did a very quick skim but did not find claims about lower risk during bull markets. In fact, my quick skim probably resulted in me misunderstanding some of the research, so take everything armed with several saltshakers.

For what it's worth, I've read suggestions to rebalance only very infrequently, such as every few years, from I think Bernstein, based on the idea of picking times far enough apart that the autocorrelation is low. The rebalancing timelines considered here are quite a bit more frequent than that.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

lack_ey wrote:For what it's worth, I've read suggestions to rebalance only very infrequently, such as every few years, from I think Bernstein, based on the idea of picking times far enough apart that the autocorrelation is low. The rebalancing timelines considered here are quite a bit more frequent than that.
You could do that, but someone who rebalanced at the end of every even year could have a fairly different allocation from someone who rebalanced at the end every odd year. Infrequent rebalancing adds randomness to the asset allocation, depending on the rebalancing dates chosen. I will again suggest partial rebalancing -- identify the trades that would get you back to your target allocation, but only execute a fraction of them.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

However the article says rebalancing increases risks in both up and down markets. I'm hoping for an explanation of how risk is increased in an up market when stocks are sold and bonds are bought
From the paper abstract:
Rebalancing is similar to starting with a buy and hold portfolio and adding a short straddle (selling both a call
and a put option) on the relative value of the portfolio assets. The option-like payoff to
rebalancing induces negative convexity by magnifying drawdowns when there are pronounced
divergences in asset returns.
The paper defines risk as negative convexity, which is the increasing probability of relative losses as a function of the divergence in performance between stocks and bonds. Most commonly this risk involves exacerbating the magnitude of portfolio drawdowns due to rebalancing out of bonds into stocks during equity bear markets, when there is a negative stock vs. bond divergence. However, it also applies in a lesser degree to positive stock - bond divergences. Primarily this is due to the fact that you are pulling money out of the positively trending asset (stocks) and reducing portfolio return.

Not discussed is my view that there is higher portfolio volatility during equity bear markets and lower volatility during equity bull markets, but I think this is implicit. If this is correct, a fixed allocation rebalancer is actually increasing portfolio volatility by rebalancing into stocks during bear markets and is magnifying the reduction of portfolio volatility during equity bull market periods by rebalancing into bonds. If he wanted to maintain a more constant level of portfolio volatility over bull and bear market periods, one way to do that is to not rebalance at all. Another way to do it is to try to use the author's "momentum overlay" technique, in which you use a trend-following method, such as moving averages, to try to compensate for the increased risk of constant-mix rebalancing. For example, in an equity drawdown you'd go ahead and rebalance as usual, but you would reduce the amount that you moved from bonds to stocks according to your "momentum overlay" alogorithm. The authors prefer this to not rebalancing at all, since it systematically avoids a large degree of "allocation drift" over time.
We don't know where we are, or where we're going -- but we're making good time.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Harvey paper says re-balancing adds risk

Post by Kevin M »

Browser wrote:Stocks are actually more risky when they're going down, not less risky.
I don't know when stocks are going down. I only know that they have gone down. I don't know what they'll do tomorrow, or even the next minute.

Knowing that stocks are going down implies that I know that they will continue to go down. If I could know when stocks will continue to go down, I could make a lot more money.

Haven't read the article, but it seems that it's based on the premise that there is a persistent "momentum factor", and that one can profit by tilting toward it. As pointed out, less frequent rebalancing is another way to do so. I'm skeptical of any technique that relies on technical analysis, and any mention of downtrends or uptrends implies that it does.

I know that there is research that seems to support that momentum is a risk factor. I'm already skeptical enough about my tilts to small and value, and am not ready to add any more tilts to my portfolio.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

Kevin M wrote:
Browser wrote:Stocks are actually more risky when they're going down, not less risky.
I don't know when stocks are going down. I only know that they have gone down. I don't know what they'll do tomorrow, or even the next minute.
Yes, but empirically, stocks are more volatile in the period *following* a negative return than after a positive return. Researchers have developed asymmetric GARCH models and stochastic volatility models with a negative correlation between the return and volatility innovations to account for this effect.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Harvey paper says re-balancing adds risk

Post by Kevin M »

Browser wrote:Risk (aka volatility) is not constant over time, so rebalancing to a fixed policy allocation doesn't make sense and never made sense as a "risk management" technique:
I think that for most investors the word "volatility" implies relatively large, short-term price movements. My investing textbooks don't define risk as volatility, but as uncertainty of return over the investor's holding period, and that definition sits well with me. Since my intended holding period for stocks generally is years, not days or months, I'm not so concerned with uncertainty of daily or monthly returns, therefore I don't equate daily or monthly volatility with risk.

I agree that stock prices are more volatile when investors are afraid, and investors typically are more afraid after stock prices have gone down, probably proportionate to how fast and by how much they've gone down. This is reflected in the VIX volatility index, which I don't pay attention to, but which I hear about now and then when there is fear in the market.

I'm not saying that stocks are not risky long term, or that perhaps even large decreases in overall stock prices don't reflect increased risk, but just that short-term volatility is not risk in my view. It may be for others, but not for me.

With respect to rebalancing being a risk-management technique, perhaps it's better to describe it as a risk/expected-return management technique. So maybe stocks are riskier after having gone down a lot, but it also may be the case that valuations are then cheaper, and longer-term expected returns are higher. And the converse may be true as well. So some sort of rebalancing seems consistent with maintaining some sort of target risk/expected-return balance.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

rmelvey wrote:
letsgobobby wrote:when stocks are moving up, rebalancing decreases risk (moving stocks to bonds). When stocks are moving down, rebalancing adds risk (moving bonds to stocks). As pointed out upthread, the goal is to maintain risk at a fairly constant level.
You are not getting to the heart of the idea though because you are comparing apples and oranges (stocks and bonds). To understand why it adds risk you have to examine the dynamic of rebalancing between two assets with equal risk.
letsgobobby does get to the heart of the matter -- it is the portfolio risk that matters -- apples AND oranges.

Billy
Easy Rhino
Posts: 3278
Joined: Sun Aug 05, 2007 11:13 am
Location: San Diego

Re: Harvey paper says re-balancing adds risk

Post by Easy Rhino »

so to truly reduce risk, only rebalance on the way up and not on the way down?
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Harvey paper says re-balancing adds risk

Post by Aptenodytes »

Risk is the wrong term, I think. A lot of the confusion expressed in the posts here reflect that. The point of the paper is that rebalancing with momentum increases return, relative to generic rebalancing, by both lowering the downward fall during declining markets and by raising the upward climb during rising markets.

The term risk makes sense with respect to lowering the downward fall during declining markets, because one way of thinking about risk is the depth of the bottoms. But on the upside it doesn't really make sense. The article is about using momentum to goose your returns.

I'll point out that paper has absolutely zero actionable information in it. To be actionable it would have to include the effect of the cost of the momentum strategy, which one can guess is not cheap. Their graphs assume that the firms they work for provide the momentum strategy for free. Perhaps someone should take them up on the implied offer.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Harvey paper says re-balancing adds risk

Post by JoMoney »

The idea of rebalancing to maintain a "risk profile" at a constant percentage of assets makes sense and his worked out quite well for decades, probably millennia (e.g. Talmud portfolio), we have at least one mutual fund that's been following a balanced stock/bond strategy since 1929.

This constant-mix percentage / rebalanced portfolio is a good empiricly proven strategy, on theoretical grounds as well laid out in things like Modern Portfolio Theory and Kelly Critierion... but I still don't like the idea of forced periodic rebalancing based on a constant-mix percentage. I really don't like the often promulgated idea of "rebalancing" for essentially market-timing strategies that believe you can outperform through rebalancing.
The risks that matter to me are tied closer to fixed dollar amounts, not percentages of assets. The only strategy that can work out well for everyone is buy and hold according to their needs and preferences, in aggregate the sum result of everyone rebalancing doesn't change the markets aggregate portfolio. There is no aggregate benefit or "bonus" for one person playing a momentum game and someone else playing a mean-reversion game.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Harvey paper says re-balancing adds risk

Post by Aptenodytes »

Holy smokes. My earlier post was based on the paper at SSRN. I just read the MarketWatch post and was amazed. It is 100% wrong.

The SSRN paper he claims to be summarizing is arguing that adding a momentum "overlay" improves rebalancing by nudging returns upwards.

The MarketWatch piece says that rebalancing is more risky than not rebalancing, but every single illustration in the paper shows that this is simply not the case. E.g.

Image

In each graph, over the time periods graphed, rebalancing and non-rebalancing perform about equally well, with rebalancing coming out a tiny bit ahead.

There is only one very narrow sense in which the column's thesis is correct, but this caveat is only mentioned at the end:
If you are a risk-averse shorter-term investor, he added, you need to be especially aware of rebalancing’s consequences.

Even this isn't quite right, because it isn't that short-term investors need to be "especially" aware, these are the only investors that need to be aware, because over periods of more than about two years there is simply no effect.

So here's the message: if you are the kind of person who is investing for a less than two-year time horizon, with on the order of 60% stocks, don't bother rebalancing. What makes it clear that this is journalistic malpractice is that this is not the most relevant advice to give such an investor. The most relevant advice would be "Why in the world is a risk-averse short-term investor holding 60% stocks?" Telling such a person to add a momentum overlay is bonkers.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

Aptenodytes wrote:I'll point out that paper has absolutely zero actionable information in it. To be actionable it would have to include the effect of the cost of the momentum strategy, which one can guess is not cheap. Their graphs assume that the firms they work for provide the momentum strategy for free. Perhaps someone should take them up on the implied offer.
If the momentum overlay tends to trade in the opposite direction as the rebalancing, it is possible that addding the momentum overlay reduces turnover and transaction costs. Skimming the paper, I did not see a comparison of turnover for rebalancing with and without a momentum overlay.
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Harvey paper says re-balancing adds risk

Post by Aptenodytes »

Beliavsky wrote:
Aptenodytes wrote:I'll point out that paper has absolutely zero actionable information in it. To be actionable it would have to include the effect of the cost of the momentum strategy, which one can guess is not cheap. Their graphs assume that the firms they work for provide the momentum strategy for free. Perhaps someone should take them up on the implied offer.
If the momentum overlay tends to trade in the opposite direction as the rebalancing, it is possible that addding the momentum overlay reduces turnover and transaction costs. Skimming the paper, I did not see a comparison of turnover for rebalancing with and without a momentum overlay.
But how do you know what to do when? Don't you have pay their firms to tell you? Or can you and I do it with freely available information and algorithms? I thought all the momentum strategies can with significant costs.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Harvey paper says re-balancing adds risk

Post by JoMoney »

Aptenodytes wrote:But how do you know what to do when? Don't you have pay their firms to tell you? Or can you and I do it with freely available information and algorithms? I thought all the momentum strategies can with significant costs.
If someone was taking this idea to the extreme you could probably just choose to invest more of your assets as prices rise. If you started with a 50/50 portfolio and the market moved it to 60/40 you'd maybe double-down and put it at 70/30. You could probably work out something using a "relative strength indicator" as well.
It shouldn't be any more expensive then someone using a mean-reversion strategy. One of them is selling their "risk" the other is buying it. Aggregate result = more transaction fees for both.

Or one could just "buy and hold" and it will just naturally have momentum like tendencies.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Harvey paper says re-balancing adds risk

Post by in_reality »

Khanmots wrote:
in_reality wrote:I make no claim to knowing what is best but ...

Intuitively to me it has always seemed that rebalancing into stocks after a crash is definitely adding risk. That's it's likely to pay off eventually is another thing...
Rebalancing is about managing risk.
Yes, it is about managing risk. Your view is just too simplistic for me.
Khanmots wrote:Say you were at 50/50 equity/FI and that matched your need and ability to take risk.
The market crashes. You're now at 25/75. Your portfolio is now less risky than it was. You re-balance back to 50/50. Yes, you added risk... but only back to the point had previously decided was necessary and right.
If you keep doing this in a falling market though, your total lifetime contributions to equities will be higher than the 50/50 you set to match your need and ability to take risk.

Also, as others have pointed out it's not like you can equate the risk of a 50/50 portfolio in normal times to the risk of a 50/50 portfolio in turbulent times. Stocks do become more risky so to rebalance back to the original allocation I don't believe puts me in the same place. This is especially true when employment stability is factored in and you are in an industry that will cut on a dime.

I don't buy the author's solution particularly since it's not easy to implement and could be costly.

If every downturn was like 2008, then I'd have no trouble rebalancing. It's the possibility of a more protracted decline that makes me hesitant to rebalance without a complete reanalysis of my ability and need to take risk.

OK I used to talk to executives in the Japanese financial industry quite a bit and perhaps they have mis-colored my thinking.

Still, increasing my total lifetime contributions in riskier assets can only mean that the effect of the rebalancing is to effectively increase the amount of risk that I am taking. To me it's not possible to view it any other way. Again, whether or not it is prudent to take that increased risk is another question - and I may do so in the future. It clearly is taking more risk though.
livesoft
Posts: 86075
Joined: Thu Mar 01, 2007 7:00 pm

Re: Harvey paper says re-balancing adds risk

Post by livesoft »

I read the article and the paper. My take-away is Do not rebalance based on the calendar. And try to Rebalance when momentum changes.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
NightOwl
Posts: 664
Joined: Fri Feb 06, 2009 1:08 am
Location: New York, NY

Re: Harvey paper says re-balancing adds risk

Post by NightOwl »

I've read a lot of threads about rebalancing -- doesn't it always come down to whether we're in a mean reversion world or a momentum world? Do we ever know which one we're in? Do we pay Mr. Harvey to assume we're in a momentum world and act accordingly?

NightOwl
"Volatility provokes the constant dread that some investors know more than we do, making us fearful of ignoring such powerful price movements." | Peter Bernstein, "The 60/40 Solution."
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

NightOwl wrote:I've read a lot of threads about rebalancing -- doesn't it always come down to whether we're in a mean reversion world or a momentum world? Do we ever know which one we're in?NightOwl
It's not just mean reversion vs. momentum. There is overwhelming evidence that the stock market is more volatile after falls than rises. The 2nd moment (variance) has some predictability, but the evidence for predictability of the 1st moment (the mean) is tenuous. Smoothing the rebalancing in bear markets as I have suggested could make sense even if you think there is no predictability in the mean.
Topic Author
VennData
Posts: 580
Joined: Mon Feb 26, 2007 4:52 pm

Re: Harvey paper says re-balancing adds risk

Post by VennData »

Livesoft sums it up nicely; So I will change my IPS re-balance rule from:
Sell in May and stay that way.
To:
Buy when things are here to stay.
in_reality, any thoughts on this subject? I'm trying to get the whole IPS into the form of an easy-to-remember Haiku
Work hard, save today
Buy when things are here to stay
Withdraw four percent
magneto
Posts: 1060
Joined: Sun Dec 19, 2010 9:57 am
Location: On Chesil Beach

Re: Harvey paper says re-balancing adds risk

Post by magneto »

Find the theory around rebalancing confusing, and this thread enlightening in the conflict between value and momentum. Presumably the momentum investor will be buying as asset class rises, while the value investor will be selling? Definitely raises a dilemma about rebalancing?

Long ago in pre BH days came to the conclusion that Process Control, where there is a desired value or set point (target AA%), and measured value (current allocation), could teach us something where response to changes and stability is concerned. Now written into our IPS (now we know what an IPS is, many thanks!), as follows :-

"Method for Rebalancing Asset Allocations
Buying as a market falls feels like throwing good money after bad, watching with dismay as ever more money goes down the drain. On the other hand selling as the market rises leads to regret as the investor is left on the sidelines watching helplessly as the market grinds relentlessly ever upwards. The aim of the method is to avoid excessive actions and retain psychological balance.

Rebalancing is commonly based on time or deviation bands. This method attemps to use both in a progressive manner and borrows from classical control theory (integral action), to seek stability.

At regular intervals, say every month or quarter, look at the deviation from the desired AA and rebalance 10% of the way to the desired AA. This response restrains the rate of approach to the desired AA, captures some momentum effect, and is a calming influence. The investor is always moving in the right direction, towards, but seldom reaching the desired AA, and thus not making large changes which could lead to regret. The method could be sub-optimal, but suit those of a cautious nature.

For Stocks the desired AA could be either a fixed target or a dynamic target based on valuations, we choose a dynamic target responding to valuations ...... "
Unquote

Thought I would chuck this in as relating loosely to the momentum v rebalancing dilemma.

Thanks for a great thread.
'There is a tide in the affairs of men ...', Brutus (Market Timer)
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

It's important to define portfolio risk, or the type of portfolio risk, that portfolio rebalancing is intended to address. Here are three definitions of risk in relation to portfolio rebalancing:

Risk as Portfolio Volatility: Rebalancing helps control the level of portfolio volatility; i.e. rebalanced portfolios will have a smaller range of volatility over time than non-rebalanced portfolios with the same starting allocation.

However, starting in 1972 with a mix of 50% TSM / 50% 5-yearTreasury notes, during every time period from 1 year to 29 years, the annualized SD of an unrebalanced portfolio was lower or equal to that of an annually rebalanced portfolio. No benefit in "risk control" for rebalancing whatsoever over these time periods beginning in 1972 and ending in 1973 through 2000. Starting in 2000 with a 50/50 mix, the annualized SD for the unrebalanced portfolio was significantly less than that of the annually rebalanced portfolio for every period from 1 year to 14 years. The effect of rebalancing on the volatility of your portfolio depends on the specific time period, and particularly on the volatility of stocks during that time period. If the volatility of stocks was constant over time, then the effect of rebalancing would be predictable. But it isn't, so the effect on portfolio volatility is unpredictable.

Risk as deviation from the Market Portfolio: William Sharpe defines risk in terms of deviation from the "market portfolio" - which is the average AA held by all investors in the aggregate.

For simplicity, assume the market portfolio at Point A is 50% stocks and 50% bonds. An investor who holds 50/50 is holding a portfolio with the average risk level. Let's assume that stocks lose 50% while bonds gain 10%. Now the market portfolio has become 31.25% stocks and 68.75% bonds. An investor who does not rebalance continues the hold a portfolio with the average risk level in terms of the market portfolio; i.e., his risk level has not changed. An investor who does rebalance to 50/50 is now holding a portfolio with an above average risk level; i.e, he has increased the risk of his portfolio. He is a contrarian who is taking on a 50% stock allocation, even though the "average" market allocation to stocks is now only 31%. Sharpe does not advocate rebalancing, because he sees risk in terms of deviation from the market portfolio.

Risk as Negative Convexity Risk is defined as the magnitude of drawdowns of overall portfolio returns as a function of the divergence between asset returns. This is the risk being addressed in this thread.
Rebalancing is similar to starting with a buy and hold portfolio and adding a short straddle (selling both a call and a put option) on the relative value of the portfolio assets. The option-like payoff to rebalancing induces negative convexity by magnifying drawdowns when there are pronounced divergences in asset returns.
The returns of a non-rebalanced portfolio, by definition, does not have negative convexity and avoids this kind of risk.
We don't know where we are, or where we're going -- but we're making good time.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

magneto wrote: At regular intervals, say every month or quarter, look at the deviation from the desired AA and rebalance 10% of the way to the desired AA. This response restrains the rate of approach to the desired AA, captures some momentum effect, and is a calming influence. The investor is always moving in the right direction, towards, but seldom reaching the desired AA, and thus not making large changes which could lead to regret. The method could be sub-optimal, but suit those of a cautious nature.
This is a fuller elaboration of what I wrote upthread:
Beliavsky wrote:My simple, free suggestion is to rebalance gradually. Do 10% of the projected rebalancing trade each month, instead of doing the entire trade immediately.
livesoft
Posts: 86075
Joined: Thu Mar 01, 2007 7:00 pm

Re: Harvey paper says re-balancing adds risk

Post by livesoft »

VennData wrote:Livesoft sums it up nicely; So I will change my IPS re-balance rule from:
[…]
in_reality, any thoughts on this subject? I'm trying to get the whole IPS into the form of an easy-to-remember Haiku
My thoughts have often been expressed: Check to see if you need to rebalance when something has an RBD. I think this helps one buy on a possible downside momentum reversal. But how to signal a reversal of upside momentum?

if one looks at emerging markets index fund (VEIEX) it will have lost money over the past 12 months by days end, but in the meantime one could have made more than 23% by buying on or around the RBD and selling at the high.
Wiki This signature message sponsored by sscritic: Learn to fish.
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

I have a question. Maybe there's a simple answer but it's escaping me now.

Most people agree that it makes sense to rebalance to maintain AA when adding "new money" to a portfolio. Please correct me if I'm wrong that "most people" advise doing this.

What is the theoretical or practical basis why using new money to rebalance should be treated differently from rebalancing through buying/selling existing assets (other than the expenses of buying/selling)? The money itself doesn't know where it came from, old or new.

Billy
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Harvey paper says re-balancing adds risk

Post by Beliavsky »

BillyG wrote:I have a question. Maybe there's a simple answer but it's escaping me now.

Most people agree that it makes sense to rebalance to maintain AA when adding "new money" to a portfolio. Please correct me if I'm wrong that "most people" advise doing this.

What is the theoretical or practical basis why using new money to rebalance should be treated differently from rebalancing through buying/selling existing assets (other than the expenses of buying/selling)? The money itself doesn't know where it came from, old or new.
If you have $400K in bonds and $200K in stocks and are supposed to get back to 50-50 (you are bond-heavy because stocks have fallen), moving $100K of bonds to stocks means that you are reducing your amount of money in safe assets, at a time when stocks are more volatile than usual. As I and others have said, stocks are more volatile after they have fallen. At least allocating new money to stocks from earned income does not reduce your position in safe assets.
leonard
Posts: 5993
Joined: Wed Feb 21, 2007 10:56 am

Re: Harvey paper says re-balancing adds risk

Post by leonard »

It does both.

If I rebalance and buy bonds, I've decreased risk.

If I rebalance and buy stocks, I've increased risk.

Pretty straightforward.
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Harvey paper says re-balancing adds risk

Post by tadamsmar »

Kevin M wrote:
Browser wrote:Stocks are actually more risky when they're going down, not less risky.
I don't know when stocks are going down. I only know that they have gone down. I don't know what they'll do tomorrow, or even the next minute.
+1

To paraphrase Ayn Rand, check your premises for misleading verb tenses.
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

Beliavsky wrote:
BillyG wrote:I have a question. Maybe there's a simple answer but it's escaping me now.

Most people agree that it makes sense to rebalance to maintain AA when adding "new money" to a portfolio. Please correct me if I'm wrong that "most people" advise doing this.

What is the theoretical or practical basis why using new money to rebalance should be treated differently from rebalancing through buying/selling existing assets (other than the expenses of buying/selling)? The money itself doesn't know where it came from, old or new.
If you have $400K in bonds and $200K in stocks and are supposed to get back to 50-50 (you are bond-heavy because stocks have fallen), moving $100K of bonds to stocks means that you are reducing your amount of money in safe assets, at a time when stocks are more volatile than usual. As I and others have said, stocks are more volatile after they have fallen. At least allocating new money to stocks from earned income does not reduce your position in safe assets.
But if I have $100K of new money to invest it is in cash, which also is a safe asset. Why is that different from moving $100K from bonds?

Billy
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

I have yet to hear a good answer as to why the different treatment for rebalancing with new money versus buy/sell existing funds.
Billy
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Harvey paper says re-balancing adds risk

Post by JoMoney »

BillyG wrote:I have yet to hear a good answer as to why the different treatment for rebalancing with new money versus buy/sell existing funds.
Billy
As far as I'm concerned, you answered the question yourself. There's an expense when buying/selling in both fees and taxes, so unless your portfolio moves way out of proportion relative to your desired target allocation, it makes sense to keep it in balance by using new money rather than buying and selling.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Harvey paper says re-balancing adds risk

Post by Aptenodytes »

BillyG wrote:I have yet to hear a good answer as to why the different treatment for rebalancing with new money versus buy/sell existing funds.
Billy
I think for most people it is a question of how frequently to rebalance. Most people use some kind of bands to avoid constantly rebalancing. New money goes in according to the AA; old money is subject to the bands.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

Probably the best thing most people could do is to forget to rebalance. After a few years they'd wake up with more money than if they had slavishly followed this mantra and be happy they forgot to do it. Stay unbalanced -- I like you just the way you are. :beer
We don't know where we are, or where we're going -- but we're making good time.
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

JoMoney wrote:
BillyG wrote:I have yet to hear a good answer as to why the different treatment for rebalancing with new money versus buy/sell existing funds.
Billy
As far as I'm concerned, you answered the question yourself. There's an expense when buying/selling in both fees and taxes, so unless your portfolio moves way out of proportion relative to your desired target allocation, it makes sense to keep it in balance by using new money rather than buying and selling.
So it's just the expenses that are different? That is what I thought although people seem to treat new money and rebalancing differently.

I can still rebalance through my tax-sheltered accounts so there are no other expenses (no buy/sell expenses for these funds).

If I have a 60/40 stock/bond allocation are there some who say new money should be allocated in this fashion (60/40) and avoid any rebalancing until there is a different trigger such as bands or calendar?

Perhaps there is not a big difference when adding new money if the existing investments are at 62/38, for example.

Billy
User avatar
DonCamillo
Posts: 1050
Joined: Tue Nov 26, 2013 9:27 pm
Location: Northern New Jersey

Re: Harvey paper says re-balancing adds risk

Post by DonCamillo »

Browser wrote:Probably the best thing most people could do is to forget to rebalance. After a few years they'd wake up with more money than if they had slavishly followed this mantra and be happy they forgot to do it. Stay unbalanced -- I like you just the way you are. :beer
It is hard to avoid at least partial rebalancing with new money if you have a defined asset allocation.

I agree with an earlier comment that rebalancing is a bet on mean reversion. But even if mean reversion does not occur, the effects of rebalancing can be hidden by inflation and GDP growth. I suspect that what most Bogleheads want is either to preserve or grow capital in real after-tax terms. Taxes and inflation can confuse most of our conclusions on these forums.
Les vieillards aiment à donner de bons préceptes, pour se consoler de n'être plus en état de donner de mauvais exemples. | (François, duc de La Rochefoucauld, maxim 93)
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Harvey paper says re-balancing adds risk

Post by Browser »

DonCamillo wrote:
Browser wrote:Probably the best thing most people could do is to forget to rebalance. After a few years they'd wake up with more money than if they had slavishly followed this mantra and be happy they forgot to do it. Stay unbalanced -- I like you just the way you are. :beer
It is hard to avoid at least partial rebalancing with new money if you have a defined asset allocation.

I agree with an earlier comment that rebalancing is a bet on mean reversion. But even if mean reversion does not occur, the effects of rebalancing can be hidden by inflation and GDP growth. I suspect that what most Bogleheads want is either to preserve or grow capital in real after-tax terms. Taxes and inflation can confuse most of our conclusions on these forums.
That seems like a good point. For example, if I have my monthly contributions going 50/50 to stocks and bonds then I would be "partially rebalancing" back to 50/50 all the time even if I did nothing else. That might be good enough for me, without shifting assets. The only other thing I might do is shift assets if my allocation really got out of whack. But all the little percentage here and little percentage there stuff is probably ineffective and not worth fussing about, IMO.
We don't know where we are, or where we're going -- but we're making good time.
BillyG
Posts: 427
Joined: Sat Nov 17, 2012 8:02 pm
Location: DC, USA

Re: Harvey paper says re-balancing adds risk

Post by BillyG »

Browser,

That sounds about right. It probably doesn't make much difference except to Bogleheads who overthink and overanalyze on their way to "staying the course." Nothing wrong with that.

And you're right that forgetting to rebalance sometimes ends up well in terms of letting momentum work do its thing.

Billy
Johno
Posts: 1883
Joined: Sat May 24, 2014 4:14 pm

Re: Harvey paper says re-balancing adds risk

Post by Johno »

leonard wrote:It does both.

If I rebalance and buy bonds, I've decreased risk.

If I rebalance and buy stocks, I've increased risk.

Pretty straightforward.
True but not necessarily equally, unless the risk of stocks is constant. But as Beliavsky mentioned, it's also pretty straightforward intuitively that by doing nothing in the midst of a stock market rout you've already increased your risk at least in the near time. Back to the options terminology quoted in the initial post, at-the-money puts on stock index futures will have gotten a lot more expensive in the midst of a rout. That refers especially to short term options, but the long term is just a series of 'short terms'. One opinion is that taking on higher risk by buying into stock downturns is more than compensated; some research challenges this or finds that it's conditional (on waiting for momentum to play out, and so forth). But of course all attempts to answer such questions through historical data have the same basic weakness that the future doesn't have to replicate the past.

My previous suggestion about rebalancing in down markets is to perhaps simply sell index puts rather than go long additional underlying. Say there's a certain core unleveraged position in underlying stock (mutual funds, ETF's, or long position in stock index futures fully backed by cash). Then if the market drops enough to push the allocation down through a specified barrier, add to stock allocation by selling the same notional amount of at-the-money index futures puts you'd otherwise buy in the underlying (IOW 1 put on the ES contract rather than 1 ES contract, same notional though less 'delta'). Say they're 1 month puts (which are liquid); then roll them (new strike always a-t-m) as long as or until the market goes back up and indicates a reduction in stock position, in which case allow them to roll off without replacement. One might sell at-the-money calls if the market goes up enough to indicate selling some of the core position. In this method you get compensated automatically for the market's changing perception of risk as expressed in options prices (which isn't infallible, but 'efficient market' never means 'infallible market', same goes for how the market prices the underlying), in regard to your periodic change in positions.

Also back to historical data, selling 1 one month index puts as a way to invest generally has beaten investing in the underlying, notional for notional no net leverage, in not only Sharpe ratio but actual total return (a rather remarkable result, it should have worse total return because of the cushioned downside and limited upside of receiving the premium) over 1986-2011 period (IIRC, about that period) per Whaley's data collected for the CME. In simple terms this result indicates the market tends to overprice puts across the cycle of ups and downs, not an unfavorable result for a strategy involving selling puts, even though the reason for the strategy is to calibrate risk rather than make excess return per se. And again no such historical anomaly is gteed to be replicated in the future.
Post Reply