Risk, Return, and Rebalancing
Risk, Return, and Rebalancing
Something I'm mulling over is the relationship between risk and rebalancing. The conventional notion is that rebalancing helps to control risk, where risk is viewed as magnitude of point in time drawdown, volatility or standard deviation of periodic (e.g., daily, monthly, annual returns). By maintaining a fixed asset mix via rebalancing, one hopes to maintain a relatively constant drawdown or variability of periodic portfolio returns. What is troubling me is the definition of risk that is being assumed in this argument.
Specifically, what we are really interested in is the compound returns produced by our investment portfolio -- not the periodic returns. If I'm invested in something, I'm concerned with the final cumulative returns produced by that investment. Given this definition, risk should be viewed in terms of the risk of the variability of cumulative returns over a given investment horizon. For example, let's say that I invest a lump sum into a portfolio that is 50% stocks and 50% bonds. I intend to hold this portfolio for 20 years. Let us say that at the end of five years, my stock allocation has grown to 60% of my portfolio. Should I now rebalance back to 50%? From the conventional perspective, the answer would be yes because a 60/40 allocation would have a higher likely variability of periodic returns than a 50/50 allocation.
But from a cumulative return perspective, should I rebalance? Well, that would have to be evaluated from the perspective of the variability of compound returns going forward. Assuming my original investment horizon hasn't changed, that would suggest that I take a look at the variability of 15-year compound returns for a 60/40 portfolio compared to the variability of the 15-year compound returns for a 50/50 portfolio. Undoubtedly, the variability of both, based on 15-year compound returns, would be greater than the 20-year variability of the 50/50 allocation that was my starting point. So, actually, in order to maintain the compound return risk level I started with I should rebalance to a stock allocation that is less than the initial 50%; perhaps now it ought to be 45%.
However, we need to consider rebalancing in terms of the variability of the dollar amount of cumulative returns and not just in terms of the variability of return percentage. It is obvious that by systematically holding the allocation percentage constant we are forfeiting some of the expected dollar amount returns. For example, we are likely to find that an initial 50/50 allocation that is rebalanced periodically will have a lower range of 20-year cumulative returns than a 50/50 portfolio that is never rebalanced, but it will also have a lower expected minimum and maximum return. The dollar-denominated risk (defined as the minimum-maximum range of returns) is actually greater than the dollar-denominated risk of a non-rebalanced portfolio.
From the above perspective, why should I ever rebalance except to attempt to manage the dollar-denominated range of cumulative portfolio returns over a given investment horizon? As the horizon becomes shorter, I would rebalance based on that shorter horizon in order to target the same expected range of expected dollar-denominated cumulative returns? Doesn't this approach make more sense?
Specifically, what we are really interested in is the compound returns produced by our investment portfolio -- not the periodic returns. If I'm invested in something, I'm concerned with the final cumulative returns produced by that investment. Given this definition, risk should be viewed in terms of the risk of the variability of cumulative returns over a given investment horizon. For example, let's say that I invest a lump sum into a portfolio that is 50% stocks and 50% bonds. I intend to hold this portfolio for 20 years. Let us say that at the end of five years, my stock allocation has grown to 60% of my portfolio. Should I now rebalance back to 50%? From the conventional perspective, the answer would be yes because a 60/40 allocation would have a higher likely variability of periodic returns than a 50/50 allocation.
But from a cumulative return perspective, should I rebalance? Well, that would have to be evaluated from the perspective of the variability of compound returns going forward. Assuming my original investment horizon hasn't changed, that would suggest that I take a look at the variability of 15-year compound returns for a 60/40 portfolio compared to the variability of the 15-year compound returns for a 50/50 portfolio. Undoubtedly, the variability of both, based on 15-year compound returns, would be greater than the 20-year variability of the 50/50 allocation that was my starting point. So, actually, in order to maintain the compound return risk level I started with I should rebalance to a stock allocation that is less than the initial 50%; perhaps now it ought to be 45%.
However, we need to consider rebalancing in terms of the variability of the dollar amount of cumulative returns and not just in terms of the variability of return percentage. It is obvious that by systematically holding the allocation percentage constant we are forfeiting some of the expected dollar amount returns. For example, we are likely to find that an initial 50/50 allocation that is rebalanced periodically will have a lower range of 20-year cumulative returns than a 50/50 portfolio that is never rebalanced, but it will also have a lower expected minimum and maximum return. The dollar-denominated risk (defined as the minimum-maximum range of returns) is actually greater than the dollar-denominated risk of a non-rebalanced portfolio.
From the above perspective, why should I ever rebalance except to attempt to manage the dollar-denominated range of cumulative portfolio returns over a given investment horizon? As the horizon becomes shorter, I would rebalance based on that shorter horizon in order to target the same expected range of expected dollar-denominated cumulative returns? Doesn't this approach make more sense?
On the internet, nobody knows you're a dog.
Re: Risk, Return, and Rebalancing
It sounds to me like you think investment returns are like savings account interest. Your returns can be negative. I rebalance to insure I keep some profit and so I can avoid behavioural mistakes (selling at a low point because of panic)
Start reading the wiki to learn more. Do the work necessary to understand. You are not the first with the idea presented.
Start reading the wiki to learn more. Do the work necessary to understand. You are not the first with the idea presented.
Re: Risk, Return, and Rebalancing
The OP is absolutely correct.
Re: Risk, Return, and Rebalancing
No. Risk is variability. It means that whatever you are measuring can be more or less than than what you predicted/calculated. When you think of "drawdown" you are thinking only of the risk on the downside and that gets you confused as to the reason for rebalncing.CULater wrote:By maintaining a fixed asset mix via rebalancing, one hopes to maintain a relatively constant drawdown or variability of periodic portfolio returns.
In Sharpe ratio the riskless asset is usually thought of as 4 week T-bills. If you invest in the riskless asset you will in no way earn less than you expected but your drawdown will be lousy. Guaranteed.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
Re: Risk, Return, and Rebalancing
The OP states quite clearly that he/she considers risk to be the indetermination in total return accumulated at a future date (or total capital value, or average annualized return, which are the same).
The OP is correct that shortening the time horizon will increase that indetermination at a fixed future date (in relative terms) and therefore the logic consequence would be to continuously reduce the stocks component and shorten the bonds maturity until on D-day one is left with only zero maturity bonds (aka "cash") and no indetermination.
The OP is correct that shortening the time horizon will increase that indetermination at a fixed future date (in relative terms) and therefore the logic consequence would be to continuously reduce the stocks component and shorten the bonds maturity until on D-day one is left with only zero maturity bonds (aka "cash") and no indetermination.
-
- Posts: 2000
- Joined: Tue Feb 23, 2016 3:24 am
Re: Risk, Return, and Rebalancing
Yeah rebalancing and the AA model is concerned with controlling the percentage risk at a given time point - but you buy milk with dollars and not percent of assets and utility of extra money is not constant regardless of amount
G.E. Box "All models are wrong, but some are useful."
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
From a returns perspective, rebalancing has historically either made virtually no difference or had a significant negative effect.
For instance, a rebalanced portfolio with 60% TSM and 40% ITT starting with nothing and getting annual contributions of $10k from 1972 until now would be worth $14.21 million, but with no rebalancing, it would be worth $14.47 million.
But with other asset classes, the effect has been quite different. In the above circumstance, if we change TSM to SCV, the portfolio would be worth $25.6 million with rebalancing but $39.1 million with no rebalancing, a huge cumulative difference.
It's true that rebalancing helps to keep a portfolio's volatility, as measured by std. dev., in check, but it does so over the long-term by selling the consistently outperforming stocks to buy the 'safe' bonds.
For instance, a rebalanced portfolio with 60% TSM and 40% ITT starting with nothing and getting annual contributions of $10k from 1972 until now would be worth $14.21 million, but with no rebalancing, it would be worth $14.47 million.
But with other asset classes, the effect has been quite different. In the above circumstance, if we change TSM to SCV, the portfolio would be worth $25.6 million with rebalancing but $39.1 million with no rebalancing, a huge cumulative difference.
It's true that rebalancing helps to keep a portfolio's volatility, as measured by std. dev., in check, but it does so over the long-term by selling the consistently outperforming stocks to buy the 'safe' bonds.
The Sensible Steward
-
- Posts: 2000
- Joined: Tue Feb 23, 2016 3:24 am
Re: Risk, Return, and Rebalancing
Consider standard deviation of cash in nominal terms without withdrawals as perhaps the limit of your argument.willthrill81 wrote:From a returns perspective, rebalancing has historically either made virtually no difference or had a significant negative effect.
For instance, a rebalanced portfolio with 60% TSM and 40% ITT starting with nothing and getting annual contributions of $10k from 1972 until now would be worth $14.21 million, but with no rebalancing, it would be worth $14.47 million.
But with other asset classes, the effect has been quite different. In the above circumstance, if we change TSM to SCV, the portfolio would be worth $25.6 million with rebalancing but $39.1 million with no rebalancing, a huge cumulative difference.
It's true that rebalancing helps to keep a portfolio's volatility, as measured by std. dev., in check, but it does so over the long-term by selling the consistently outperforming stocks to buy the 'safe' bonds.
G.E. Box "All models are wrong, but some are useful."
Re: Risk, Return, and Rebalancing
So many definitions of risk out there.
One is the risk of running out of money, or not having the money when needed.
From a longer term real world (rather than academic volatility) perspective, this cautious investor prefers looking at upside potential v downside risk, which will fluctuate with the markets.
Good post OP !!!
One is the risk of running out of money, or not having the money when needed.
From a longer term real world (rather than academic volatility) perspective, this cautious investor prefers looking at upside potential v downside risk, which will fluctuate with the markets.
Good post OP !!!
'There is a tide in the affairs of men ...', Brutus (Market Timer)
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
I really don't understand the logic of rebalancing for someone who won't need to withdraw from a portfolio for 15 or more years. To me, it makes more sense to determine how much cash one will need in the next 10 years, and put that into some low volatility assets. The rest of the portfolio would be in equities. Once a decumulation period starts, it might make sense to start swapping equities for bonds, but only in order to meet the anticipated 10 year needs.
One could argue that a so-called balanced portfolio outperformed equities over the three decades after 1982. But isn't that long bond bull market due to the long and slow decline in interest rates over that period? Does anyone really believe that such a decline will ever happen again? I certainly do not.
One could argue that a so-called balanced portfolio outperformed equities over the three decades after 1982. But isn't that long bond bull market due to the long and slow decline in interest rates over that period? Does anyone really believe that such a decline will ever happen again? I certainly do not.
-
- Posts: 577
- Joined: Fri Mar 15, 2013 8:49 pm
Re: Risk, Return, and Rebalancing
Anything is possible...the future is cloudy....etc etc etc. My take is that data from the past is unreliable no matter how you look at it. All that data is dependent on what happened to happen during that period of human history. There's no law that says the last hundred years has to look like the next. Things change and evolve in ways that are completely unpredictable, and you only get one ride on this merry go round. Best to hedge your bets.North Texas Cajun wrote:I really don't understand the logic of rebalancing for someone who won't need to withdraw from a portfolio for 15 or more years. To me, it makes more sense to determine how much cash one will need in the next 10 years, and put that into some low volatility assets. The rest of the portfolio would be in equities. Once a decumulation period starts, it might make sense to start swapping equities for bonds, but only in order to meet the anticipated 10 year needs.
One could argue that a so-called balanced portfolio outperformed equities over the three decades after 1982. But isn't that long bond bull market due to the long and slow decline in interest rates over that period? Does anyone really believe that such a decline will ever happen again? I certainly do not.
Re: Risk, Return, and Rebalancing
Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
http://www.norstad.org/finance/risk-and-time.html
Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Cantpassagain: "Anything is possible..."
We're talking about whether to rebalance right now. My argument is that the steady 30 year decline of yields - from 15% to 2% - cannot be repeated when bond yields are currently under 3%. So a repeat of 1982 to 2012 is not going to happen until interest rates go up a lot.
A 50/50 portfolio is not a hedge against uncertainty for an investor who doesn't need the cash for 20 years. Rather, it's a bet that equities will not outperform bonds over 20 years, as they have for the past century. It's also a bet that such a "balanced" portfolio will keep pace with inflation. That's a poor bet, IMO.
We're talking about whether to rebalance right now. My argument is that the steady 30 year decline of yields - from 15% to 2% - cannot be repeated when bond yields are currently under 3%. So a repeat of 1982 to 2012 is not going to happen until interest rates go up a lot.
A 50/50 portfolio is not a hedge against uncertainty for an investor who doesn't need the cash for 20 years. Rather, it's a bet that equities will not outperform bonds over 20 years, as they have for the past century. It's also a bet that such a "balanced" portfolio will keep pace with inflation. That's a poor bet, IMO.
-
- Posts: 2000
- Joined: Tue Feb 23, 2016 3:24 am
Re: Risk, Return, and Rebalancing
There seems to be a false dichotomy of rebalancing at a fixed AA or 100 percent equities. There should to be some more reasonable approach to shifts to fixed income taking into account income effects (both positive and negative) with changes in valuations and total absolute amounts invested. I believe in fixed income in some manner decreasing volatility in some manner. But I do not understand AA and naive rebalancing. A glide path but one that takes into account your personal values and the randomness of the stock market or a preset shifting pattern of pulling funds to fix dollars amounts of FI.pkcrafter wrote:Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
Thank you
QJ
G.E. Box "All models are wrong, but some are useful."
-
- Posts: 577
- Joined: Fri Mar 15, 2013 8:49 pm
Re: Risk, Return, and Rebalancing
That and your idea to hold 10 years of needs in low volatility investments sounds good in theory but most folks don't even have 10 years of needs saved in total. So how does one execute while still in the accumulation phase?North Texas Cajun wrote:Cantpassagain: "Anything is possible..."
We're talking about whether to rebalance right now. My argument is that the steady 30 year decline of yields - from 15% to 2% - cannot be repeated when bond yields are currently under 3%. So a repeat of 1982 to 2012 is not going to happen until interest rates go up a lot.
A 50/50 portfolio is not a hedge against uncertainty for an investor who doesn't need the cash for 20 years. Rather, it's a bet that equities will not outperform bonds over 20 years, as they have for the past century. It's also a bet that such a "balanced" portfolio will keep pace with inflation. That's a poor bet, IMO.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Pkcrafter,
I don't have time to read that entire paper. But I did see that he was comparing equity returns with holding cash - with the assumption that a bank cash account could earn 6%. How can such a comparison be relevant to any financial decision being made today?
I don't have time to read that entire paper. But I did see that he was comparing equity returns with holding cash - with the assumption that a bank cash account could earn 6%. How can such a comparison be relevant to any financial decision being made today?
-
- Posts: 2000
- Joined: Tue Feb 23, 2016 3:24 am
Re: Risk, Return, and Rebalancing
viewtopic.php?f=10&t=126809&start=50#p1862535CantPassAgain wrote:That and your idea to hold 10 years of needs in low volatility investments sounds good in theory but most folks don't even have 10 years of needs saved in total. So how does one execute while still in the accumulation phase?North Texas Cajun wrote:Cantpassagain: "Anything is possible..."
We're talking about whether to rebalance right now. My argument is that the steady 30 year decline of yields - from 15% to 2% - cannot be repeated when bond yields are currently under 3%. So a repeat of 1982 to 2012 is not going to happen until interest rates go up a lot.
A 50/50 portfolio is not a hedge against uncertainty for an investor who doesn't need the cash for 20 years. Rather, it's a bet that equities will not outperform bonds over 20 years, as they have for the past century. It's also a bet that such a "balanced" portfolio will keep pace with inflation. That's a poor bet, IMO.
G.E. Box "All models are wrong, but some are useful."
-
- Posts: 3589
- Joined: Fri Dec 26, 2014 3:19 pm
Re: Risk, Return, and Rebalancing
What happens in the next 10 years? And the 10 years after that? Are you moving a year's worth or a month's worth at a time from the "equity" bucket to the "bond" bucket? Or do you spend down the whole 10 years and then make a big move? If you are constantly refilling the "safe" bucket, then you have decided to spend from equities. If you aren't, you are spending just from bonds, right?I really don't understand the logic of rebalancing for someone who won't need to withdraw from a portfolio for 15 or more years. To me, it makes more sense to determine how much cash one will need in the next 10 years, and put that into some low volatility assets. The rest of the portfolio would be in equities. Once a decumulation period starts, it might make sense to start swapping equities for bonds, but only in order to meet the anticipated 10 year needs.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Cantpassagain: "but most folks don't even have 10 years of needs saved in total. "
Oh, but for almost every two income couple, the bulk of non-discretionary retirement needs should be met by Social Security benefits. Furthermore, for investors under 45, there should be little need to withdraw from a retirement portfolio for another 20 years.
Yes, someone who is 55 with less than $100K in savings is probably screwed. My only advice to such people would be to plan on working to age 72 and to start saving 15% of their income.
Oh, but for almost every two income couple, the bulk of non-discretionary retirement needs should be met by Social Security benefits. Furthermore, for investors under 45, there should be little need to withdraw from a retirement portfolio for another 20 years.
Yes, someone who is 55 with less than $100K in savings is probably screwed. My only advice to such people would be to plan on working to age 72 and to start saving 15% of their income.
-
- Posts: 6993
- Joined: Mon Jan 03, 2011 8:40 am
Re: Risk, Return, and Rebalancing
As much as I have read and thought about the topic I don't really care for ANY Of the financial definition of risk. In the end risk is you don't have X dollars when you need to have X dollars. It is up to each individual to figure out how to do that. If that means LMP great. If that means pensions and SS and annuities great. If that means heavy stocks forever great.
In the end there is no one answer. Trust me I am sure MANY on here wish it was that simple.
Good luck
In the end there is no one answer. Trust me I am sure MANY on here wish it was that simple.
Good luck
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” |
-Jack Bogle
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
notwhoyouthink wrote: "If you are constantly refilling the "safe" bucket, then you have decided to spend from equities. If you aren't, you are spending just from bonds, right?"
I think that depends on a lot of factors. I expect to refill my safe bucket after equities rise. In some years I will spend from my safe bucket and in others I will spend from my equity bucket. So I will deplete my safe bucket at times, but never 100%. I simply do not believe equities will stay down for 10 years. If I am wrong, my standard of living may drop slightly.
IMO, bonds are nowhere near as safe as some people believe. Bonds have looked good for the past three decades. But that's only because of the very high interest rates of the early 1980s, and the fact that those rates took 30 years to fully decline back to historical levels.
I think that depends on a lot of factors. I expect to refill my safe bucket after equities rise. In some years I will spend from my safe bucket and in others I will spend from my equity bucket. So I will deplete my safe bucket at times, but never 100%. I simply do not believe equities will stay down for 10 years. If I am wrong, my standard of living may drop slightly.
IMO, bonds are nowhere near as safe as some people believe. Bonds have looked good for the past three decades. But that's only because of the very high interest rates of the early 1980s, and the fact that those rates took 30 years to fully decline back to historical levels.
- bertilak
- Posts: 10711
- Joined: Tue Aug 02, 2011 5:23 pm
- Location: East of the Pecos, West of the Mississippi
Re: Risk, Return, and Rebalancing
If you cut off that chart at 20 years, will the final std deviation numbers be any different?pkcrafter wrote:Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
P.S. I think that the OP is correct. As your planning horizon gets shorter you should lower your risk. That is the whole point of the "age in bonds" and similar rules of thumb. But I would not put things in terms of re-balancing but re-allocating. I would re-balance to whatever allocation I thought was currently appropriate. So yes, re-balancing IS a matter of restoring risk levels but that is a moving target.
I think the OP and I are saying the same thing but with a slightly different emphasis. Of course the same AA has a different risk for a different planning horizon, but perhaps things get out out of balance balance faster than the planning horizon changes. There is a great deal of fuzziness (and unknowable-ness) about the actual ideal AA to begin with.
May neither drought nor rain nor blizzard disturb the joy juice in your gizzard. -- Squire Omar Barker (aka S.O.B.), the Cowboy Poet
Re: Risk, Return, and Rebalancing
But you would not be rebalancing in the hope of making more money (in fact you will likely make less), but with the intent of making your final result less uncertain!North Texas Cajun wrote:Cantpassagain: "Anything is possible..."
We're talking about whether to rebalance right now. My argument is that the steady 30 year decline of yields - from 15% to 2% - cannot be repeated when bond yields are currently under 3%. So a repeat of 1982 to 2012 is not going to happen until interest rates go up a lot.
A 50/50 portfolio is not a hedge against uncertainty for an investor who doesn't need the cash for 20 years. Rather, it's a bet that equities will not outperform bonds over 20 years, as they have for the past century. It's also a bet that such a "balanced" portfolio will keep pace with inflation. That's a poor bet, IMO.
You are absolutely correct that bonds are not gonna yield as much as they did in the past, that they are not so safe as some assume, and let me add that their correlation to stocks will be higher than many expect.
Yet, as maturity gets shorter bonds become safer and the safest investment vehicle there is are shorter-term bonds.
-
- Posts: 18461
- Joined: Tue Dec 31, 2013 6:05 am
- Location: 26 miles, 385 yards west of Copley Square
Re: Risk, Return, and Rebalancing
My understanding is that stocks and bonds are seen as inversely correlated. If one goes way up, you see your AA going out of a band and rebalance. If both diverge a lot, you see your AA go out of band and rebalance. The expectation is that the class that's gone up will come down and the class that's gone down will come up. Reversion to the mean. This occurs even if both are rising but one is rising much faster. So it's reducing volatility and increasing return. I'd add *in theory* to all of this.
So how often would one rebalance. My understanding is that Vanguard does it daily in it's target date funds. I do it yearly. Others say to do it no more often than quarterly. I don't know the answer.
So how often would one rebalance. My understanding is that Vanguard does it daily in it's target date funds. I do it yearly. Others say to do it no more often than quarterly. I don't know the answer.
Bogle: Smart Beta is stupid
Re: Risk, Return, and Rebalancing
They are actually seen as uncorrelated (historically): if one goes up, the other can equally well go up, go down, or do nothing.Jack FFR1846 wrote:My understanding is that stocks and bonds are seen as inversely correlated.
We live in interesting times, though. The overwhelmingly major driver of bond prices at present are central banks' monetary policies. Less directly, they also affect stock prices and arguably are a major driver for those too.
This means correlation between those two assets is as high as it has ever been.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Jack FFR1886 wrote: "My understanding is that stocks and bonds are seen as inversely correlated."
The data I've seen show that U.S. equities and long term treasuries were inversely correlated from 1970 to 2000. However, that correlation reversed after 2000, and the two asset classes have been positively correlated.
So which period should we use as a predictor for the next 20 years?
The data I've seen show that U.S. equities and long term treasuries were inversely correlated from 1970 to 2000. However, that correlation reversed after 2000, and the two asset classes have been positively correlated.
So which period should we use as a predictor for the next 20 years?
Re: Risk, Return, and Rebalancing
Not sure which correlation you guys are looking at.
This is the historical rolling 3-year correlation, which is generally oscillating around a slightly negative value.
This is the historical rolling 3-year correlation, which is generally oscillating around a slightly negative value.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Thesaints: "This is the historical rolling 3-year correlation, which is generally oscillating around a slightly negative value."
That's not exactly what I'm seeing. It appears to me that the correlation between stocks and treasuries oscillated around zero prior to 1970. From 1970 to 2000, the two asset classes were negatively correlated. From 2000 to the end of the chart they were positively correlated.
I don't think we can conclude that correlation will be negative or even slightly negative over the next 20 years, which I think is the planning horizon for the OP.
That's not exactly what I'm seeing. It appears to me that the correlation between stocks and treasuries oscillated around zero prior to 1970. From 1970 to 2000, the two asset classes were negatively correlated. From 2000 to the end of the chart they were positively correlated.
I don't think we can conclude that correlation will be negative or even slightly negative over the next 20 years, which I think is the planning horizon for the OP.
Last edited by North Texas Cajun on Wed Jul 19, 2017 3:14 pm, edited 1 time in total.
Re: Risk, Return, and Rebalancing
This correlation question is highly susceptibility to the question being asked.Thesaints wrote:Not sure which correlation you guys are looking at.
This is the historical rolling 3-year correlation, which is generally oscillating around a slightly negative value.
Are we addressing bonds in general or Treasuries in particular?
Are we talking of a 100 years of market history or the last six months in 2008?
The correlation can be essentially non existent or highly negative depending on the question.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Doc: "The correlation can be essentially non existent or highly negative depending on the question."
Well, if we use the chart provided by Thesaints, wecan see that the correlation between U.S. stocks and U.S. treasuries has been highly positive for the past 16 years. This recent trend seems long enough to indicate that U.S. treasuries may no longer provide the diversification benefit that some once believed they did.
I agree with the argument that treasuries provide near-term safety. But that's not the same role as an asset which provides diversification.
Doc, I'm not meaning to challenge your statement. I really would like to know if there is a fixed income asset class which in recent decades has had a highly negative correlation with U.S. equities.
Well, if we use the chart provided by Thesaints, wecan see that the correlation between U.S. stocks and U.S. treasuries has been highly positive for the past 16 years. This recent trend seems long enough to indicate that U.S. treasuries may no longer provide the diversification benefit that some once believed they did.
I agree with the argument that treasuries provide near-term safety. But that's not the same role as an asset which provides diversification.
Doc, I'm not meaning to challenge your statement. I really would like to know if there is a fixed income asset class which in recent decades has had a highly negative correlation with U.S. equities.
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
I think that one's take on this issue depends greatly on what purpose you are holding stocks and bonds for. Traditionally, the view is that stocks are for growth and bonds are for safety, and the historical data support this. In retirement, my plan is to generally hold ten years of spending in bonds, leaving the remainder in stocks. This is a significant departure from the traditional AA strategies that use percentages, but this 'bucketed' approach makes more intuitive sense to me. My plan is to spend the money from stocks when they are up, but when stocks are down to spend from the bond money until the stocks recover and the ten year spending in bonds can be replenished. I haven't yet figured out exactly what metric I would use to determine when to move funds from stocks to the bonds in this situation though.
As such, the question then becomes when to start building this spending supply in bonds. As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark.
Using this approach, there is no place for rebalancing until I'm in retirement and need to move money from recovered equities into bonds.
As such, the question then becomes when to start building this spending supply in bonds. As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark.
Using this approach, there is no place for rebalancing until I'm in retirement and need to move money from recovered equities into bonds.
The Sensible Steward
- patrick013
- Posts: 3301
- Joined: Mon Jul 13, 2015 7:49 pm
Re: Risk, Return, and Rebalancing
CULater wrote:From the above perspective, why should I ever rebalance except to attempt to manage the dollar-denominated range of cumulative portfolio returns over a given investment horizon? As the horizon becomes shorter, I would rebalance based on that shorter horizon in order to target the same expected range of expected dollar-denominated cumulative returns? Doesn't this approach make more sense?
Does history repeat itself ? You don't need CAPE10 to figure this out.
Both spots with arrows are when the standard PE hit 25 and above. Even
with a 50-50 portfolio you're going to be spending quite a few bonds for
a number of years - if history repeats itself.
Living with volatility with a long horizon...age-in-bonds starts to look better.
age in bonds, buy-and-hold, 10 year business cycle
Re: Risk, Return, and Rebalancing
Depends on your time frame. Treasuries generally have a high negative correlation to equities in a flight to quality situation like a stock market crash or sometimes even only in a correction.North Texas Cajun wrote:I really would like to know if there is a fixed income asset class which in recent decades has had a highly negative correlation with U.S. equities.
Three month rolling return chart: Intermediate Treasuries and S&P 500
http://quotes.morningstar.com/chart/fun ... %22%3A3%7D
It would not surprise me that if you did a regression on this data for the total time period you would find only a small negative correlation if any at all. But if you look at many of the shorter time periods of the chart you get a different story. Just look at the areas where the orange drops below zero. What does the blue do then?
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Thesaints: "But you would not be rebalancing in the hope of making more money (in fact you will likely make less), but with the intent of making your final result less uncertain!"
I agree that, over a 20 year period, an 80/20 or 100/0 portfolio would be "less uncertain" than a 50/50 portfolio. But the uncertainty is almost all on the upside. Over a 20 year period it is unlikely that an 80/20 portfolio will underperform a 50/50 portfolio. When it does underperform, the amount it underperforms will not be that great (again, over 20 years). On the other hand, the upside potential is much larger.
Now, I am not saying that an investor should not be in bonds. I think bonds do provide safety, and are appropriate for near term (0 to 7 or even 0 to 10 years of non-discretionary spending). But why rebalance a retirement portfolio if one is 20 years from retirement? And why rebalance any more than is necessry to meet 10 year non-discretionary requirements?
Of course, one may believe that the correlation between equities and bonds will reverse the 16 year trend and once again be negative. In that case, perhaps one can improve returns by rebalancing.
I agree that, over a 20 year period, an 80/20 or 100/0 portfolio would be "less uncertain" than a 50/50 portfolio. But the uncertainty is almost all on the upside. Over a 20 year period it is unlikely that an 80/20 portfolio will underperform a 50/50 portfolio. When it does underperform, the amount it underperforms will not be that great (again, over 20 years). On the other hand, the upside potential is much larger.
Now, I am not saying that an investor should not be in bonds. I think bonds do provide safety, and are appropriate for near term (0 to 7 or even 0 to 10 years of non-discretionary spending). But why rebalance a retirement portfolio if one is 20 years from retirement? And why rebalance any more than is necessry to meet 10 year non-discretionary requirements?
Of course, one may believe that the correlation between equities and bonds will reverse the 16 year trend and once again be negative. In that case, perhaps one can improve returns by rebalancing.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Willthrill81: "As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark."
Willthrill81, that's not exactly what I did, but probably what I would do if I could repeat that period (ten years before retirement). I think it's a smart plan.
One clarification, though: my bond allocation is now my ten years of spending requirements less the fixed income we will receive from social security and pensions.
Willthrill81, that's not exactly what I did, but probably what I would do if I could repeat that period (ten years before retirement). I think it's a smart plan.
One clarification, though: my bond allocation is now my ten years of spending requirements less the fixed income we will receive from social security and pensions.
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Doc wrote: "It would not surprise me that if you did a regression on this data for the total time period you would find only a small negative correlation if any at all. But if you look at many of the shorter time periods of the chart you get a different story. Just look at the areas where the orange drops below zero."
Thanks for the chart and the thoughts.
I guess the very short term negative correlation is why I think of treasuries (and other bonds) as appropriate for short term (0 to 7 year) spending requirements. But probably why I do not see bonds as much help for long term diversification.
Thanks for the chart and the thoughts.
I guess the very short term negative correlation is why I think of treasuries (and other bonds) as appropriate for short term (0 to 7 year) spending requirements. But probably why I do not see bonds as much help for long term diversification.
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
Thanks. This approach just makes more sense to me than arbitrarily selecting something like a 60/40 AA. Historically, ten years of safe spending is enough cushion to get through even the worst bear markets with the portfolio intact. And considering that I'll reach SS age around fifteen years into retirement, that will add another margin of safety, never mind the fact that we could cut back on our discretionary spending and extend the ten years of bond money into closer to fifteen.North Texas Cajun wrote:Willthrill81: "As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark."
Willthrill81, that's not exactly what I did, but probably what I would do if I could repeat that period (ten years before retirement). I think it's a smart plan.
One clarification, though: my bond allocation is now my ten years of spending requirements less the fixed income we will receive from social security and pensions.
Once I reach 70 or whatever the maximum age for SS is by the time I get there, my amount in bonds will certainly drop since my needed portfolio income will drop by the amount of my SS payouts, though I'll probably still maintain the ten years of spending.
The Sensible Steward
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Patrick013 wrote: "Both spots with arrows are when the standard PE hit 25 and above."
Not sure what you are implying here. The S&P 500 trailing 12 month PE was above 25 from Feb 1998 to Sep 2003.
http://www.multpl.com/table?f=m
Not sure what you are implying here. The S&P 500 trailing 12 month PE was above 25 from Feb 1998 to Sep 2003.
http://www.multpl.com/table?f=m
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
The correlation between monthly returns in TSM and TBM over the last 25 years is .01, essentially zero. If you look at daily returns, the correlation is -.16, which is still very small.North Texas Cajun wrote:Doc wrote: "It would not surprise me that if you did a regression on this data for the total time period you would find only a small negative correlation if any at all. But if you look at many of the shorter time periods of the chart you get a different story. Just look at the areas where the orange drops below zero."
Thanks for the chart and the thoughts.
I guess the very short term negative correlation is why I think of treasuries (and other bonds) as appropriate for short term (0 to 7 year) spending requirements. But probably why I do not see bonds as much help for long term diversification.
The Sensible Steward
Re: Risk, Return, and Rebalancing
Long ago, I stopped going out of my way to make investing more confusing.
The benefit does not outweigh the headache.
JT
The benefit does not outweigh the headache.
JT
Re: Risk, Return, and Rebalancing
It depends on what you are going to do when the stock market tanks and it will several times in an investing lifetime. Are you going to sell bonds and buy equities or just say thank you Doc that I have some bond to keep my total portfolio form dropping even further.North Texas Cajun wrote:Doc wrote: "It would not surprise me that if you did a regression on this data for the total time period you would find only a small negative correlation if any at all. But if you look at many of the shorter time periods of the chart you get a different story. Just look at the areas where the orange drops below zero."
Thanks for the chart and the thoughts.
I guess the very short term negative correlation is why I think of treasuries (and other bonds) as appropriate for short term (0 to 7 year) spending requirements. But probably why I do not see bonds as much help for long term diversification.
FWIW Diversification in not a goal in itself it is just a means for attaining a better overall portfolio. Each of us may have a different definition of better but that is an individual risk/reward choice that we must choose depending on our own needs and (possibly) fortitude.
To paraphrase Larry Swedroe our AA is based on our own individual ability, need and willingness to take risk. The Norstad link that pkcrafter provided up thread has some interesting ideas related to this. http://www.norstad.org/finance/risk-and-time.html
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
Re: Risk, Return, and Rebalancing
Something I've struggled with is what happens if you get to the ten years out from retirement date and your stock funds are off, say, 20%? Do you still move a year's worth of spending into bonds?willthrill81 wrote:I think that one's take on this issue depends greatly on what purpose you are holding stocks and bonds for. Traditionally, the view is that stocks are for growth and bonds are for safety, and the historical data support this. In retirement, my plan is to generally hold ten years of spending in bonds, leaving the remainder in stocks. This is a significant departure from the traditional AA strategies that use percentages, but this 'bucketed' approach makes more intuitive sense to me. My plan is to spend the money from stocks when they are up, but when stocks are down to spend from the bond money until the stocks recover and the ten year spending in bonds can be replenished. I haven't yet figured out exactly what metric I would use to determine when to move funds from stocks to the bonds in this situation though.
As such, the question then becomes when to start building this spending supply in bonds. As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark.
Using this approach, there is no place for rebalancing until I'm in retirement and need to move money from recovered equities into bonds.
By starting ten years out maybe the issue is resolved because you've built in enough wiggle room.
But one thing I always wonder about when I read posts like these is what you do if you're at a very down time at the time when you want to be starting to keep your portfolio safer.
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
That's a good point, and in the event of a bear market like that, i would probably just delay moving that money into bonds. If I was 'forced' to enter retirement with 'only' 8 years of spending in bonds, I'd still be comfortable.MoonOrb wrote:Something I've struggled with is what happens if you get to the ten years out from retirement date and your stock funds are off, say, 20%? Do you still move a year's worth of spending into bonds?willthrill81 wrote:I think that one's take on this issue depends greatly on what purpose you are holding stocks and bonds for. Traditionally, the view is that stocks are for growth and bonds are for safety, and the historical data support this. In retirement, my plan is to generally hold ten years of spending in bonds, leaving the remainder in stocks. This is a significant departure from the traditional AA strategies that use percentages, but this 'bucketed' approach makes more intuitive sense to me. My plan is to spend the money from stocks when they are up, but when stocks are down to spend from the bond money until the stocks recover and the ten year spending in bonds can be replenished. I haven't yet figured out exactly what metric I would use to determine when to move funds from stocks to the bonds in this situation though.
As such, the question then becomes when to start building this spending supply in bonds. As of now, I'm planning on moving one year of spending from stocks to bonds (or a nice SVF in my 401k) ten years out from projected retirement and continuing that until I reach the ten year mark.
Using this approach, there is no place for rebalancing until I'm in retirement and need to move money from recovered equities into bonds.
By starting ten years out maybe the issue is resolved because you've built in enough wiggle room.
But one thing I always wonder about when I read posts like these is what you do if you're at a very down time at the time when you want to be starting to keep your portfolio safer.
The Sensible Steward
Re: Risk, Return, and Rebalancing
Samuelson made the point a while back that equities don't become less risky as a function of time if you view it in terms of compound return; time diversification is false. But that's only true if returns are a random walk, but not true if returns exhibit mean regression. If you assume mean regression or "what goes up must go down and vice versa", (and a lot of posters seem to be assuming that, such as the "bucket theorists"), then it's more likely that compound return variability becomes smaller as a function of the length of holding period. That would imply that you should reduce stock holdings as your investment horizon becomes shorter in order to maintain the same (or lower) compound return risk level. If you don't believe in mean regression then it doesn't make any difference and there would be no reason to reduce stock holdings over time, except that you might have less need/desire to assume portfolio risk as you are aging.pkcrafter wrote:Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
On the internet, nobody knows you're a dog.
- bertilak
- Posts: 10711
- Joined: Tue Aug 02, 2011 5:23 pm
- Location: East of the Pecos, West of the Mississippi
Re: Risk, Return, and Rebalancing
Perhaps you should be easing into that position so there is no sudden disruption.MoonOrb wrote:But one thing I always wonder about when I read posts like these is what you do if you're at a very down time at the time when you want to be starting to keep your portfolio safer.
May neither drought nor rain nor blizzard disturb the joy juice in your gizzard. -- Squire Omar Barker (aka S.O.B.), the Cowboy Poet
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Doc wrote: "Are you going to sell bonds and buy equities or just say thank you Doc that I have some bond to keep my total portfolio form dropping even further.?"
Certainly not going to thank you. If you have read my posts above, you will see that, as a retiree, I am already keeping 7 to 10 years of spending in bonds. I never said that bonds do not have a place in a portfolio. I said that there is little need for rebalancing to meet some target allocation. Furthermore, I see little need for an accumulating investor to keep anything in bonds until he reaches (retirement - 10 years).
IMO, diversification into bonds for a investor such as the OP - who said he will not need the funds for 20 years - is overrated.
Certainly not going to thank you. If you have read my posts above, you will see that, as a retiree, I am already keeping 7 to 10 years of spending in bonds. I never said that bonds do not have a place in a portfolio. I said that there is little need for rebalancing to meet some target allocation. Furthermore, I see little need for an accumulating investor to keep anything in bonds until he reaches (retirement - 10 years).
IMO, diversification into bonds for a investor such as the OP - who said he will not need the funds for 20 years - is overrated.
Last edited by North Texas Cajun on Wed Jul 19, 2017 4:55 pm, edited 1 time in total.
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
The big issue I have with the argument that stocks don't become less risky over time is that the likelihood of both losing money and achieving below average returns have both absolutely decreased as the holding period increased, at least historically.CULater wrote:Samuelson made the point a while back that equities don't become less risky as a function of time if you view it in terms of compound return; time diversification is false. But that's only true if returns are a random walk, but not true if returns exhibit mean regression. If you assume mean regression or "what goes up must go down and vice versa", (and a lot of posters seem to be assuming that, such as the "bucket theorists"), then it's more likely that compound return variability becomes smaller as a function of the length of holding period. That would imply that you should reduce stock holdings as your investment horizon becomes shorter in order to maintain the same (or lower) compound return risk level. If you don't believe in mean regression then it doesn't make any difference and there would be no reason to reduce stock holdings over time, except that you might have less need/desire to assume portfolio risk as you are aging.pkcrafter wrote:Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
The Sensible Steward
- willthrill81
- Posts: 32250
- Joined: Thu Jan 26, 2017 2:17 pm
- Location: USA
- Contact:
Re: Risk, Return, and Rebalancing
The historical likelihood of bonds beating stocks over a 20 year period is extremely low. I believe it's only happened twice since 1802, and it was only true of long-term bonds, which most bond investors aren't holding.North Texas Cajun wrote:IMO, diversification into bonds for a investor such as the OP - who said he will not the funds for 20 years - is overrated.
The Sensible Steward
-
- Posts: 235
- Joined: Sun Jul 02, 2017 8:56 am
Re: Risk, Return, and Rebalancing
Moonorb: "Something I've struggled with is what happens if you get to the ten years out from retirement date and your stock funds are off, say, 20%? Do you still move a year's worth of spending into bonds?"
Almost every investor I know who is in his 50s is adding more to his retirement portfolio than at any other time of his life. That's generally because they earn more at that age, they have fewer non-discretionary spending needs, and they have more of a sense of urgency. Such an investor who finds himself in a bear market could choose to put his new savings 100% into bonds, allowing his older equity savings to return to pre-bear levels.
I'm not sure I would do that myself. In fact, I faced exactly that choice in 2008, five years before my retirement. I think my wife started to buy bonds with 50% of her new savings, while I remained investing 100% equities.
Almost every investor I know who is in his 50s is adding more to his retirement portfolio than at any other time of his life. That's generally because they earn more at that age, they have fewer non-discretionary spending needs, and they have more of a sense of urgency. Such an investor who finds himself in a bear market could choose to put his new savings 100% into bonds, allowing his older equity savings to return to pre-bear levels.
I'm not sure I would do that myself. In fact, I faced exactly that choice in 2008, five years before my retirement. I think my wife started to buy bonds with 50% of her new savings, while I remained investing 100% equities.
Re: Risk, Return, and Rebalancing
That may be true if you consider every rolling time period for U.S. stocks over the last 100+ years. But as the length of time period increases, there are fewer and fewer independent time periods as data points -- far too few to draw the conclusion you have. And if you think that longer time periods are always safe, then consider the 15 years from 1968 through 1982, when after adjusting for inflation the S&P 500 lost a total of 4.62%.willthrill81 wrote:The big issue I have with the argument that stocks don't become less risky over time is that the likelihood of both losing money and achieving below average returns have both absolutely decreased as the holding period increased, at least historically.CULater wrote:Samuelson made the point a while back that equities don't become less risky as a function of time if you view it in terms of compound return; time diversification is false. But that's only true if returns are a random walk, but not true if returns exhibit mean regression. If you assume mean regression or "what goes up must go down and vice versa", (and a lot of posters seem to be assuming that, such as the "bucket theorists"), then it's more likely that compound return variability becomes smaller as a function of the length of holding period. That would imply that you should reduce stock holdings as your investment horizon becomes shorter in order to maintain the same (or lower) compound return risk level. If you don't believe in mean regression then it doesn't make any difference and there would be no reason to reduce stock holdings over time, except that you might have less need/desire to assume portfolio risk as you are aging.pkcrafter wrote:Let's have a look at the chart on the bottom of this page on Risk and Time.
http://www.norstad.org/finance/risk-and-time.html
Paul
On the internet, nobody knows you're a dog.