1966-1982 - Stock Market return of zero.

[Cut-Throat]

We have discussed this topic before, but I have slighty different question, which I am wondering if anyone has seen a discussion on this.

During this period the market and various asset classes fluctuated quite a bit. If someone was holding a portfolio and rebalanced at the end of every year, I am guessing that he/she would have had a positive return over this period than someone that did not rebalance.

The rebalancing would have forced the portfiolio holder to sell winners and buy some losers. Has anyone ever seen a discussion or a sample portfolio with rebalancing during this period?

 

[cute fuzzy bunny]

I think that would depend on which asset classes you owned.

Both US stocks and bonds did cycle up and down during this period, but poorly and in many cases the cycling was harmonious rather than divergent. While I havent looked lately, I think real estate also stunk during that same time period. There were some opportunities in some foreign markets from time to time.

You would have had to depend on simple cycle timing - pick a high and low oscillator and buy and sell when you hit those levels. With the usual market timing luck no doubt.

Hopefully we know better than to repeat the policies that resulted in this periods behavior. If not, I hope everyone has at least a couple of mil put away to get through it.

 

[wabmester]

Has anyone ever seen a discussion or a sample portfolio with rebalancing during this period?

Doesn't FIREcalc do this for you? I just ran it with a 16-year payout period. It told me that the 16-year period starting in 1969 was the worst time to retire. I ran it with a 5% withdraw, which of course is 100% safe for such a short period, but at the end my nest egg was about 1/3 the size I started with (with annual rebalancing). Play around with the inputs.

On the subject of monetary policy mistakes, we won't know if we're making a mistake this time around until the chickens come home to roost. And this time, the experiment is on a global scale.

What happens to inflation/deflation when Japan prints money like crazy to buy US debt at a time when our treasury debt is soaring, our trade deficit is widening, and our dollar is dropping? Who knows, but nothing like it has ever been sustainable in the past.

 

[Cut-Throat]

Doesn't FIREcalc do this for you?

If I'm not mistaken, Firecalc just give you how the whole market behaves and does not rebalance into different asset classes.

What I am wondering about is say that someone had a 25x25x25x25 in Large Growth & Value and Small Growth & Value and rebalanced into that each year, how they would have fared?

 

[JWR1945]

There are numerous sources that verify the theoretical advantages of re-balancing annually. The trouble is that most studies gloss over the fees of re-balancing. If it is just a matter of where to put new money or what to sell when making withdrawals, re-balancing is free. Otherwise, you may pay more than it is worth. Generally, expect to re-balance only after you have an allocation shift of 5% to 10% if you have to pay for it.

Gummy has a theoretical study at his site. He was surprised to learn that it increased his overall return even if the other security was inferior to stocks. William Bernstein has produced several theoretical studies, but they are flawed because they assume zero cost. My preference is what raddr posted way-back-when at the www.nofeeboards.com site. I cannot imagine that the www.retireearlyhomepage.com site does not look into re-balancing as well. I have not checked, but it is likely that this site includes several references.

Have fun.

John R.

 

[unclemick]

In theory I agree with John R - except that we don't do it because of taxable and non-taxable, his and hers - and we are no longer in the 'accumulation phase'. So we hold balanced index 'within each box' and let them rebalance themselves.

I believe Bogle's (1994 book) had section which supported rebalancing vs set and forget. The library might have a copy. It's one of the books I can't keep - 4 copies 'borrowed in the last ten yrs'. My 4th purchase of Ben Graham's Intelligent Investor - I keep hidden.

As mentioned, costs matter when rebalancing and with taxable/tax deferred funds - it can get complicated.

 

[Cut-Throat]

As mentioned, costs matter when rebalancing and with taxable/tax deferred funds - it can get complicated.

I was thinking of a strategy of holding all index funds in the Tax Deferred fund. This would more or less elimiate rebalancing costs. Put a foreign fund and a REIT in the mix and I'm betting that the 1966-1982 period does not look as bad as they say.

 

[sgeeeee]

I believe FIRECALC does assume rebalancing. And I'm certain that intercst's SWR program assumes you rebalance. In fact, with SWR I think you can specify the month that you rebalance.

 

[Cut-Throat]

SG,

I believe FIRECALC does assume rebalancing

Forgive me SG, but I cannot even see where you can input different asset classes in FIRECalc.

For example Let's say you wanted a portfolio that was 40% Large Cap Growth Stocks, 20% Large Cap Value Stocks, 20% International Stocks, 20% Small Cap Value Stocks for your Stock Component for an overall 60/40 Stock/ Bond Split.

How would FIREcalc know how to rebalance to these specific asset classes?

 

[ats5g]

Cut-Throat,

If there is just one column or whatever to input "stock" returns, you can use Excel to calculate the rebalanced return of a LG/LV/SG/SV portfolio pretty easily, and then copy and past those rebalanced returns into the SWR spreadsheet.

I took returns from Gummy's site (and w/ a little help from W.Bernstein) came up with this spreadsheet. You might have to "right-click" and "save as" since geocities sometimes won't do the direct link. It calculates the nominal and real returns for the portfolios. Not very high tech. One can vary the %'s in each asset class.

Unfortunately, my "foreign stocks" data only goes back to 1970. Well, the MSCI EAFE anyway. You might try Ibbottson's SBBI (Stocks, Bonds, Bills, Inflation) for better return info.

- Alec

 

[Ted]

FIRECalc resets the asset allocation to whatever the user inputs, at the end of every historical year. This simulates rebalancing. If a person is doing it with no-load mutual funds, it is essentially cost-free, although there may be some taxes to be paid. But when you are retired, this doesn't usually matter all that much, because you will probably be in pretty much the same relatively low tax bracket whether you liquidate your taxable assets sooner or later.

In practice, retirees can accomplish most if not all rebalancing by simply withdrawing assets from the asset class that they want to reduce.

In theory, it is fine to talk about "withdrawing X% per year, adjusted for inflation." But in the real world, withdrawals are a lot more "lumpy" than that, and how many people are going to sit down and explicitly adjust their permissible spending for the coming year according to the increase in the CPI?

The main limitation on FIRECalc is that it limits a person to only two asset classes. But given that using past asset performance to project future results is a very inexact process, and that controlling annual spending is also an inexact process, FIRECalc is still very useful -- with intelligent substitution of additional asset classes for "stocks" or whatever fixed income asset is selected.

 

[Cut-Throat]

The main limitation on FIRECalc is that it limits a person to only two asset classes. But given that using past asset performance to project future results is a very inexact process

Well, my original point and post was NOT to project future results, but only as an exercise in determining what the return was from the period of 1966-1982, if a person was rebalancing yearly into let's say 5 different asset classes.

So my point was that - Maybe their return was NOT zero compared to the total market return which was zero!

 

[sgeeeee]

For example Let's say you wanted a portfolio that was 40% Large Cap Growth Stocks, 20% Large Cap Value Stocks, 20% International Stocks, 20% Small Cap Value Stocks for your Stock Component for an overall 60/40 Stock/ Bond Split.

How would FIREcalc know how to rebalance to these specific asset classes?

Sorry, Cut-Throat. I misunderstood your question. Both FIRECalc and SWR assume only two asset classes, but they do rebalance those classes. I think I recall intercst indicating that the reason he didn't offer more asset classes in the program is because he was unable to find appropriate historical data going back that far.

Of course your question would only require that you find the data for the short period of time that you want to examine. It wouldn't be too hard to develop a spreadsheet to look at your question if you had the tabulated historical data for that time period.

 

[wabmester]

Cut-Throat, I posted this link in another thread, but I'll include it here for completeness.

Our hero Bernstein has a good write-up on rebalancing theory here that appears to answer your question in the affirmative:

http://www.efficientfrontier.com/ef/996.pdf

 

[ryalmokas]

You might want to check out the many essays by a Robert B. Gordon at the Financial Sense Online site (http://www.financialsense.com/). In the archives section, under his name, are many fine essays on rebalancing. Mr Gordon is in his 80's, and shares his learnings from more than 50 years of investing. I'm not his agent, but I am a fan of his writings.

regards,
Dick Y