For the Duration

"Buy and Hold" is what we imagine if "for the duration" applies to investing. TV ads for brokerages and brain dead media writers are now advising people not to panic just because their accounts are down 50% from December 31, 2007. It "Bogles" the mind. "Stay the course", they say. But the course was unreliable and devastating in 2008. Buying and holding the indexes no matter what is a proven disaster course for most adult investors. For the past ten years through today, the total return of the low cost Vanguard S&P 500 Index Fund VFINX with all dividends reinvested  and no taxes deducted was -1.47% per year annualized or -13.72% total. For the same ten years and same conditions, the Vanguard Prime Money Market Fund VMMXX was up 3.48% annualized or 40.66% total. It's not hard to understand why some people always stay in money market funds.

 

Over the same ten years and conditions the Vanguard Ginnie Mae fund VFIIX was up 5.87% annualized and 76.66% total. One of my favorites of 2007 and until July 2008, Loomis Sayles Bond Fund LSBDX, was up only 6.18% annualized and 81.92% total after a huge fall this autumn.

 

If you had been planning in January 1999 to retire in ten years, in which of those funds would you have likely invested? For the then preceding ten years from January 6, 1989 to January 6, 1999 VFINX (SP500) was up 19.26% annualized or 484.98% total.  The money market fund was up only 5.62% annualized or 72.95% total for those past ten years. Let me guess which fund you or I might well have chosen.

 

Things change, markets change, and we need to change too. John Hussman's January 5  weekly comment  shows a way to adapt to change which is based upon an investor's age and when that investor will need to begin drawdowns for family events ranging from college expenses to retirement. As I have written many times, the worst  thing that can happen to someone expecting to retire in a year or two (or five or more) is to lose 40-50% of saved assets in a bear market year. That happened in 2002, and it happened to some folks in 2008, and it will happen again some other year.

 

Hussman shows how an annual re-balancing of investments can incorporate the concept of asset class "duration" as a way to limit over-exposure to volatile stocks which can lead to 40-50% losses. As discussed here often in evaluating bond funds, duration refers to price sensitivity to changes in interest rates. Duration for most bonds or funds is the time to when you get your total bond payment stream on a present value basis. (Read Hussman's 2004 discussion for the math.)

 

A money market fund theoretically has no price risk. It is pegged at $1 and never goes up or down, so its duration is zero. The Vanguard Short Term Municipal fund (VWSTX/VWSUX) has an average maturity of 1.2 years and an average duration of 1.1 years. One effect of this is that the bond fund will go up about 1.1% if general interest rates fall 1% or down about 1.1% if interest rates rise 1%. Compare that to the Vanguard Long Term Treasury Bond Fund (VUSTX/VUSUX) with an average maturity of 17.1 years and average duration of 11.2 years. In this case a 1% rise in general rates sends the fund down 11.2%! This is why we saw long Treasury bonds move up so far and fast in the past few months and fall so fast and far in the past week. They are ten times more volatile than the short term municipals fund. We all are more or less aware of this fact. You generally get paid a lot more to take on the ten-fold risk of a long term bond. But not this year.

 

As Hussman shows, stocks also have a duration factor. It's easier to do the math on the stock duration as it is price divided by dividend rate. A very simple way to approximate the S&P 500 duration is to use this  helpful calculator. For these purposes enter a beginning date of January and ending date of December and take the ending dividend rate. Divide that into 1 (1/.0323 for 2008), round it, and you get 31 as the duration for the S&P at year's end. This is a rough estimate but gives you some idea for how much riskier stocks are than a one year duration short term municipal fund, namely 31 times riskier. At the end of June 2007 with a dividend rate of 1.73% the duration was 58 years, so it's only one-half as risky now to own stocks as it was then. At the end of 1999 with the S&P 500 dividend rate at 1.15% the duration was 87 years! At the end of 1974 the S&P 500 duration was 18.6 years.

 

OK, so what do you do with these numbers? Let's say you will be retiring in ten years. According to Hussman you'll want to match your total portfolio to your time to retirement, namely ten years. So if the stock duration is 31 you'd divide 10 years by that duration and get 32% as the maximum allocation of your 100% portfolio for stocks. That means you'd want to have 68% in money market funds with zero duration. Or, say, 25% in stocks and 75% in a bond fund or funds with a duration of 7 years.

 

If you had rebalanced your portfolio at the end of June 2007 last year for a ten year retirement duration, you would have reduced your stock allocation to a maximum of 17% of your portfolio with 83% in money market funds, or perhaps 10% stocks and 90% in a 7 year duration bond fund. If the latter you would have lost 28% of your 10% in VFINX (S&P 500), or 2.8% overall. And you would have made 8.33% on the 90% of your 7 year duration Vanguard Intermediate Term Bond Index Fund VBIIX, or 7.5% overall. Netted out 7.5%-2.8% = +4.7% annualized over the 18 month period. That sure beats losing 28% annualized by being 100% in VFINX for that period. Not spectacular but not a retirement buster.

 

I chose an ideal starting date, and these are only rough calculations, but they show what you can do without even trying to predict what the markets will do. With a simple spreadsheet, you could do this once a year as you get a year closer to when you will need your money. Check it out and do read the Hussman articles.

 

Despite all my efforts last year, and thanks to the end of December rallies in oil and gold and bonds, I came out up 0.19% for 2008 in my retirement accounts! Not nearly as well as I would have done just being in a Hussman-like duration match portfolio or even in a money market fund. Hmmmmmmm. :)

 

 

Comments

[Joe]

Tom,

with your +0.x% last year you did better in a retirment account than 99% of all people - my guess. My result was a -16.6%. I sure hope that this result taught me the lesson I needed to learn to never ever repeat anything similar. It must be said it is my 4th year working, so we're not talking about big amounts yet. But I sure do not want to lose a single penny of my hard-earned money to my own sutpidity.
Joe

[Tom Drake]

Joe,

If you have over 30 years working ahead of you could probably put all new money into stocks for the next decade. You could buy the total world index fund VT and save yourself a lot of effort. Just set an alarm clock for ten years from now....LOL

[Jim Pitinii]

This isn't relevant, but I thought you might like to know:

The Oct 08 low for the Nekkei 225 was at least a 25 year low, probably more but that's as far back as I can get.

January 2009 is the 75th anniversary of Roosevelt's new gold price of $35.00. The halfway point is July 1971, one month before Nixon closed the gold window in August.

Jim

[Tom Drake]

Someone I've known since the 1990's glory days of the University of Colorado Long Wave Center site sent me this URL today on a possible truer model for our current collapse than 1929:

http://chronicle.com/temp/reprint.php?id=477k3d8mh2wmtpc4b6h07p4hy9z83x18

That certainly was one of the classic Kondratieff Wave "falls from top". 1873-1896

[Tom Drake]

Jim,

I found this chart on Tokyo market value to GNP from 1928 to 2000. Maybe a better long term chart will surface, but this gives a framework at least:

http://screencast.com/t/gn9UM5Ov

[Tom Drake]

For Joe and others re long term S&P 500 returns...

This is another way of thinking about and visualizing S&P 500 returns based upon the holding period. My review today of Hussman's duration concept is one way to attempt to control it for age and for "when needed". This chart shows the whole S&P 500 history since 1871.

The average total returmn with all dividends reinvested but with no allowance for fees, taxes, slippage, etc. is 9.4% per year. But there have been holding periods of up to 20 years with negative returns! For the current twenty year holding period since January 6th 1989 the total return is 8.37% or "below average".

http://screencast.com/t/QjYSrUSay

This reinforces the idea that if you are young as an investor, say under 30 years of age, you can and probably should expose your portfolio 100% to equities to try to capture the average of 9.4% annual gains and the possibility of even more. But as you get over 35-40 years old you need to start reducing that exposure to limit the possibility of a bad streak reducing your end result as I decribed earlier today.

[Joe]

Tom,

An alternative to your Husman-based duration model could be a simple trend-following model in 4 or five asset classes as described by Mebane in his paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=962461
Results are in his blog:
http://worldbeta.blogspot.com/

It's a simple model about be in a market as long as it's above the 7 or so day SMA and exit otherwise. It generates according to him about 3-4 trades per year per asset class and has not had neg. returns yet. (I know sounds like Madoff.) This is a reasonable time effort in retirement or before. I think it's worth looking at.

I know at least 80% lose on market timing - we have to keep this in mind. There are a lot of psychological traps even in simple systems.

Joe

[Tom Drake]

Joe,

There are a lot of different investment methods or models if one has the time and is "methodical" about doing them for years and years without fail. My own experience is that we change, and we sometimes go through times when we are very, very busy with work or lots of work-related travel or health or family problems which take all our attention for considerable periods of time. Or we develop new interests or move our home or change careers. We may forget about our method or lose interest in it for a long while.

I think we are better off if we rationally automate investing by putting money away every month and then annually or semi-annually spending a few hours "re-balancing" the assets. You could do that even with a moving average program or any other active program. You might make a lot of money on moving averages with stocks one year or with bonds another and need to rebalance a la Hussman. Rational rebalancing doesn't exclude an active method of investing, nor the other way around.

Bear in mind that you don't just want to accumulate money in tax-deferred accounts, but in taxable accounts also whose investment goals are different and depend upon income tax rates and capital gains rates which can change dramatically over time.

As you make money and have more of it to manage, it requires greater attention. At some time in life you may decide to have a bank trust department or other reliable money manager do it for you. But by then, at least if you have done it yourself for years, you will have a much better idea of what you want them to do for you.

So investing is a growth process like life itself. There are no guarantees, but there are logical ways to progress over time. The goal is to preserve and hopefully grow your assets so that your life is pleasant as you grow older and that it enables you to give your family some advantages as they start out in life.

This is all quite obvious and simple-minded or trite, but it needs to be respected as we go along in life. ;)

Keep in mind that I am just talking as one who has been there, and I am not an investment adviser or money manager except for myself and family. This is all my story or experience.

[Joe]

Tom,

Your advice is greatly appreciated.
Joe

[humble1]

timely link to the 1873-1896 downgrade. that is EXACTLY the model i am using and the reason i have just read the hard-to-find chronicle ("on the record": d.w. perkins) of panics during that period, as i mentioned here a while back.

one of the long cycles i consider is the kress, which ends in 2014. that would be a fine time, imho, for some kind of long wave cycles (i watch several) bottom-nesting to occur.

of course, i expect the equity low to occur sooner, perhaps in the lindsay time frame as recently discussed here.

[Recoba]

Tom;

Thanks for this piece. I found it highly informative. The concept of stocks having a duration makes alot of sense, and rebalancing based on your age is priceless.

[Tom Drake]

Since Asian markets start opening in the afternoon for AZ, west coast, and Hawaii-based investors, this interesting Japanese multi-market chart and news page might be of interest. It auto-refreshes every 90 seconds and news in English updates more frequently.

http://www.geocities.jp/real_chart_fx_sgxnikkei_dow/worldstockindex.html

The chart bands and moving averages aren't detailed, and I don't read Japanese; but the intermarket comparisons are excellent. I notice that the Russian RTS chart seems stuck. I like this page a lot. It might appeal to you east coast night owls too...:))

[humble1]

you're giving out too much homework!

B)

[Tom Drake]

I get a new PhD about every six months at Internut University....I think you do too.... alas, dear olde IU will never win a BCS bowl....(8^)