Quantifying loss due to rebalancing and market timing

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Wed Oct 31, 2018 5:43 am

In relatively volatile times we are reminded of Morningstar’s Return Gap studies:


We don’t enjoy the full returns our funds provide...

The average investor has lagged behind the average fund for the past 10 years. The reason is that, in aggregate, investors’ timing is not very good. Over the 10 years ended 2016, the average U.S. investor in diversified equity funds enjoyed a 4.36% return, even though the average diversified equity fund returned 5.15%. That’s a fair amount to give up. In fixed income, the gap was nearly as large, and that’s painful because the returns are much smaller. In bondland, we found the average investor received a 2.99% return, versus 3.72% for the average bond fund.
Combining all funds, we come up with a return for the average fund of 4.33% compared with a 3.96% return for the average investor.



Passive investors do better than active investors but still blow it...

All told, the average annualized return gap was 2.48 percentage points for actively managed funds and 1.84 points for index funds.
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

UpperNwGuy
Posts: 1243
Joined: Sun Oct 08, 2017 7:16 pm

Re: Quantifying loss due to rebalancing and market timing

Post by UpperNwGuy » Wed Oct 31, 2018 6:22 am

I invest on the first of every month. What am I doing wrong?

PFInterest
Posts: 2684
Joined: Sun Jan 08, 2017 12:25 pm

Re: Quantifying loss due to rebalancing and market timing

Post by PFInterest » Wed Oct 31, 2018 6:23 am

funny, i can usually outperform.

i get the avg returns on the way up (indexes)
and i get to TLH on the way down

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Wed Oct 31, 2018 6:26 am

Ron Scott wrote:
Wed Oct 31, 2018 5:43 am
We don’t enjoy the full returns our funds provide...
Is it because we didn't lump sum our whole portfolio in Jan 2006?

longinvest
Posts: 3109
Joined: Sat Aug 11, 2012 8:44 am

Re: Quantifying loss due to rebalancing and market timing

Post by longinvest » Wed Oct 31, 2018 7:01 am

Time-weighted (portfolio) returns and money-weighted (investor) returns aren't mathematically comparable. Studies that do it probably have a hidden agenda, such as implying that individual index investors are not good at investing, that they would better pay a financial advisor instead.

Our wiki explains the difference between the two types of returns with an example: Calculating personal returns

The wiki page also provides an easy-to-use calculator (online and downloadable) that calculates both types of returns. In particular, it provides the appropriate time-weighted (portfolio) returns to compare the performance of the investor's portfolio to a benchmark. Index investors can rest easy if their time-weighted returns match the returns of their benchmark.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic / international) stocks / domestic (nominal / inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Re: Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Wed Oct 31, 2018 8:23 am

Y'all a tough crowd, for what I thought was a interesting effort by Morningstar to actually quantify what BHers already believed.

I thought most of you decry market timing and simply assume that rebalancing does not boost long-term returns.

Change of heart out there?
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Wed Oct 31, 2018 8:45 am

Ron Scott wrote:
Wed Oct 31, 2018 8:23 am
I thought most of you decry market timing and simply assume that rebalancing does not boost long-term returns.

Change of heart out there?
No, but one must compare apples to apples, even if an apples to oranges comparison supports the preferred position.

Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Re: Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Wed Oct 31, 2018 8:55 am

rkhusky wrote:
Wed Oct 31, 2018 8:45 am
Ron Scott wrote:
Wed Oct 31, 2018 8:23 am
I thought most of you decry market timing and simply assume that rebalancing does not boost long-term returns.

Change of heart out there?
No, but one must compare apples to apples, even if an apples to oranges comparison supports the preferred position.
Seems to me the annual Morningstar Return Gap study is pretty straightforward.

So you think timing purchases and sales of a fund and regular rebalancing leads to better returns that the overall long-term performance of an underlying index fund?

What exactly do you think Morningstar gets wrong in its ongoing work?
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

PFInterest
Posts: 2684
Joined: Sun Jan 08, 2017 12:25 pm

Re: Quantifying loss due to rebalancing and market timing

Post by PFInterest » Wed Oct 31, 2018 8:56 am

Ron Scott wrote:
Wed Oct 31, 2018 8:23 am
simply assume that rebalancing does not boost long-term returns.
who said? it can.
but its not the reason to re-balance.

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Wed Oct 31, 2018 9:30 am

Ron Scott wrote:
Wed Oct 31, 2018 8:55 am
rkhusky wrote:
Wed Oct 31, 2018 8:45 am
Ron Scott wrote:
Wed Oct 31, 2018 8:23 am
I thought most of you decry market timing and simply assume that rebalancing does not boost long-term returns.

Change of heart out there?
No, but one must compare apples to apples, even if an apples to oranges comparison supports the preferred position.
Seems to me the annual Morningstar Return Gap study is pretty straightforward.

So you think timing purchases and sales of a fund and regular rebalancing leads to better returns that the overall long-term performance of an underlying index fund?

What exactly do you think Morningstar gets wrong in its ongoing work?
I didn't read the study. Perhaps you can tell us how the study handled periodic contributions or whether the study only looked at a lump sum deposited in January 2006 (or would that be January 2007?)?

Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Re: Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Thu Nov 01, 2018 7:39 am

rkhusky wrote:
Wed Oct 31, 2018 9:30 am
Ron Scott wrote:
Wed Oct 31, 2018 8:55 am
rkhusky wrote:
Wed Oct 31, 2018 8:45 am
Ron Scott wrote:
Wed Oct 31, 2018 8:23 am
I thought most of you decry market timing and simply assume that rebalancing does not boost long-term returns.

Change of heart out there?
No, but one must compare apples to apples, even if an apples to oranges comparison supports the preferred position.
Seems to me the annual Morningstar Return Gap study is pretty straightforward.

So you think timing purchases and sales of a fund and regular rebalancing leads to better returns that the overall long-term performance of an underlying index fund?

What exactly do you think Morningstar gets wrong in its ongoing work?
I didn't read the study. Perhaps you can tell us how the study handled periodic contributions or whether the study only looked at a lump sum deposited in January 2006 (or would that be January 2007?)?
Neither. This is not a simple computer game like people here play with to guess their futures. And it’s obviously not limited to contributions. What makes the gap real is investor behavior—including withdrawals due to market timing, spending, rebalancing etc. Morningstar looks at cash inflows and outflows and calculates investor returns for 1, 3, 5, and 10-year periods. They compare active and passive investor returns to the actual reported fund returns and measure the gap between actual and fund results.

There is a tendency for us to talk about actual returns as being equal to the fund’s returns less fees and taxes. Maybe we get into lump sum vs. DCA. That is naive.

Here’s a summary article and the Morningstar methodology.

https://www.nytimes.com/2018/10/12/busi ... turns.html

https://corporate.morningstar.com/us/do ... dology.pdf
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

livesoft
Posts: 62933
Joined: Thu Mar 01, 2007 8:00 pm

Re: Quantifying loss due to rebalancing and market timing

Post by livesoft » Thu Nov 01, 2018 8:00 am

Anyone (OK, almost anyone) can calculate and track the performance of their portfolio over months and years. They can compare the performance of their portfolio to the benchmarks. Then they will know if they have to even think about this return gap.

We often see folks at Bogleheads.org state that they don't need to measure themselves since they invest in a 3-fund or all index fund portfolio. Do they have a behavorial gap or a return gap? We cannot and they cannot know for sure.

Perhaps Morningstar.com should find 100 to 1000 volunteers and track the transactions of those volunteers to see if their study matches an actual set of investor behavior?
Wiki This signature message sponsored by sscritic: Learn to fish.

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Thu Nov 01, 2018 8:09 am

Ron Scott wrote:
Thu Nov 01, 2018 7:39 am
Neither. This is not a simple computer game like people here play with to guess their futures. And it’s obviously not limited to contributions. What makes the gap real is investor behavior—including withdrawals due to market timing, spending, rebalancing etc. Morningstar looks at cash inflows and outflows and calculates investor returns for 1, 3, 5, and 10-year periods. They compare active and passive investor returns to the actual reported fund returns and measure the gap between actual and fund results.

There is a tendency for us to talk about actual returns as being equal to the fund’s returns less fees and taxes. Maybe we get into lump sum vs. DCA. That is naive.

Here’s a summary article and the Morningstar methodology.

https://www.nytimes.com/2018/10/12/busi ... turns.html

https://corporate.morningstar.com/us/do ... dology.pdf
Their methodology is comparing an estimate of investor return, with all its associated buying and selling, versus the total fund return, which is equivalent to lump summing the total investment at the beginning of the sample period. That is an apples to oranges comparison because investors don't get all their money to invest at the same time.

Suppose there was no selling and investors just contributed regularly (weekly, monthly, etc). In a bull market, investor returns would still be below the fund total return, which assumes that all the contributions were made at the beginning of the period. On the other hand, investor returns would be higher than the fund total return in a bear market, for the same reason.

Rather than comparing investor return to total fund return, a better comparison would be to the return one would achieve if the initial investment was spread throughout the sample period.

And I am not sure looking at a single fund is all that accurate either. Suppose someone sold some of their large growth fund and moved into a small value fund. How would just looking at the large growth fund provide insight into that investor's return? If you are just looking at a single fund, it would be better to look at a target date fund, which is more likely to be an investor's only holding. Or sum up the returns over all mutual funds.

Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Re: Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Thu Nov 01, 2018 8:47 am

rkhusky wrote:
Thu Nov 01, 2018 8:09 am
Ron Scott wrote:
Thu Nov 01, 2018 7:39 am
Neither. This is not a simple computer game like people here play with to guess their futures. And it’s obviously not limited to contributions. What makes the gap real is investor behavior—including withdrawals due to market timing, spending, rebalancing etc. Morningstar looks at cash inflows and outflows and calculates investor returns for 1, 3, 5, and 10-year periods. They compare active and passive investor returns to the actual reported fund returns and measure the gap between actual and fund results.

There is a tendency for us to talk about actual returns as being equal to the fund’s returns less fees and taxes. Maybe we get into lump sum vs. DCA. That is naive.

Here’s a summary article and the Morningstar methodology.

https://www.nytimes.com/2018/10/12/busi ... turns.html

https://corporate.morningstar.com/us/do ... dology.pdf
Their methodology is comparing an estimate of investor return, with all its associated buying and selling, versus the total fund return, which is equivalent to lump summing the total investment at the beginning of the sample period. That is an apples to oranges comparison because investors don't get all their money to invest at the same time.

Suppose there was no selling and investors just contributed regularly (weekly, monthly, etc). In a bull market, investor returns would still be below the fund total return, which assumes that all the contributions were made at the beginning of the period. On the other hand, investor returns would be higher than the fund total return in a bear market, for the same reason.

Rather than comparing investor return to total fund return, a better comparison would be to the return one would achieve if the initial investment was spread throughout the sample period.

And I am not sure looking at a single fund is all that accurate either. Suppose someone sold some of their large growth fund and moved into a small value fund. How would just looking at the large growth fund provide insight into that investor's return? If you are just looking at a single fund, it would be better to look at a target date fund, which is more likely to be an investor's only holding. Or sum up the returns over all mutual funds.
By tracking inflows over time they do get periodic investing behaviors and they’re not looking at a single fund. And apples-and-oranges is EXACTLY that they end up concluding. The funds’ reported returns are the apples and our returns the oranges.

It’s actually a well done study with results that should not surprise BHers. If you time the market or engage in the psychological arithmetic we call rebalancing, you typically underperform. Such is life...
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Thu Nov 01, 2018 12:53 pm

Ron Scott wrote:
Thu Nov 01, 2018 8:47 am
By tracking inflows over time they do get periodic investing behaviors and they’re not looking at a single fund. And apples-and-oranges is EXACTLY that they end up concluding. The funds’ reported returns are the apples and our returns the oranges.

It’s actually a well done study with results that should not surprise BHers. If you time the market or engage in the psychological arithmetic we call rebalancing, you typically underperform. Such is life...
So, you are saying that an investor that dutifully contributes monthly, with no selling whatsoever, and ends up under-performing the single fund that he is invested in, simply because he is investing in a bull market, must be doing something wrong?

Ron Scott
Posts: 1090
Joined: Tue Apr 05, 2016 5:38 am

Re: Quantifying loss due to rebalancing and market timing

Post by Ron Scott » Thu Nov 01, 2018 3:55 pm

rkhusky wrote:
Thu Nov 01, 2018 12:53 pm
Ron Scott wrote:
Thu Nov 01, 2018 8:47 am
By tracking inflows over time they do get periodic investing behaviors and they’re not looking at a single fund. And apples-and-oranges is EXACTLY that they end up concluding. The funds’ reported returns are the apples and our returns the oranges.

It’s actually a well done study with results that should not surprise BHers. If you time the market or engage in the psychological arithmetic we call rebalancing, you typically underperform. Such is life...
So, you are saying that an investor that dutifully contributes monthly, with no selling whatsoever, and ends up under-performing the single fund that he is invested in, simply because he is investing in a bull market, must be doing something wrong?
Please read posts and any relevant attachments before responding to them.

First, I am simply presenting some widely quoted research by Morningstar that demonstrates how average investors underperform fund results. The research speaks for itself as to is quality and value.

Second, neither I nor the research I attached talks about the results of a "dutiful" investor making monthly contributions sans withdrawals, but it is obviously possible that he might experience returns differing from the fund's especially during the short term.

Finally, I myself commented about timing the market and rebalancing as being the 2 behaviors of concern, not investing on a paycheck basis...

Good luck!
Retirement is a game best played by those prepared for more volatility in the future than has been seen in the past. The solution is not to predict investment losses but to prepare for them.

livesoft
Posts: 62933
Joined: Thu Mar 01, 2007 8:00 pm

Re: Quantifying loss due to rebalancing and market timing

Post by livesoft » Thu Nov 01, 2018 4:21 pm

I would not be surprised if my behavior with tax-loss harvesting shows that I am selling low most of the time instead of buy-and-hold.
Wiki This signature message sponsored by sscritic: Learn to fish.

lukestuckenhymer
Posts: 84
Joined: Wed May 30, 2018 11:53 am

Re: Quantifying loss due to rebalancing and market timing

Post by lukestuckenhymer » Thu Nov 01, 2018 4:40 pm

What's done is done. How will this information help you?

Don't market time and don't worry. You're doing your best.

https://www.youtube.com/watch?v=ZVA-Bx4rNc0

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Quantifying loss due to rebalancing and market timing

Post by rkhusky » Thu Nov 01, 2018 7:52 pm

Ron Scott wrote:
Thu Nov 01, 2018 3:55 pm
First, I am simply presenting some widely quoted research by Morningstar that demonstrates how average investors underperform fund results. The research speaks for itself as to is quality and value.
It is pretty much useless, doesn't really say anything about the returns of real investors, and provides no separation between "good" and "bad" investing techniques.

And, if you don't rebalance, then your portfolio risk is not constant, which is another apples vs oranges comparison. Talking about return, while disregarding risk, is also pretty much useless.

Post Reply