Revisit asset allocation in present environment

[Sensei]

Hi,

I'll stay with my thesis that stocks are the better generators of wealth over the 'long run'.

However, I'd also say that I agree with the mechanics suggested by the previous two posts. When I increase my bond allocation, I will be doing three things:
1. Buying the actual bonds, and not a fund. Even the lowest MERs I've seen are too high.
2. Building a portfolio of bonds (a bond ladder) that meets my income needs when I buy the bonds and not relying on capital appreciation. IOW, a port from which I can live off the interest income, although in my case it will most definitely be combined with dividend income.
3. This goes with 2. planning on holding the bonds to maturity.

[big easy]

You could also reduce the duration of your bonds by limiting yourself to 1-5 yr bonds instead of 1-10yr. This would reduced volatility while perserving capital.

[Justise]

I still want to see the math. Yes, a bond will lose market value when interest rates are rising. But that bond also creeps towards maturity year over year, causing its price to converge towards its face value regardless of rate environment; and new units purchased at that point go in at the higher yield. In a bond fund, the higher-duration bonds will tick down over those years and converge towards their own face values.

There's a huge difference between bond prices that drop because of interest rates, and share prices of AIG/Citigroup/GE that drop due to investor confidence. Shares of those companies have no underlying redemption value. Bonds do, regardless of the market prices they encounter along the way, and duration is what measures how far their market prices will deviate. If I hold a 20-year government bond and the market value drops 20% due to a rise in interest rates, it doesn't mean there's a risk the bond will go bankrupt. I will still collect all the payment and recoup the face value if I hold it to maturity.

The math is simple, in Canadian market context; there is a practically complete cycle of rising interest rate of environment of 31 years from 1950 to 1981 and a falling interest rate environment of 29 years from 1981 to 2010, though both the rise and fall are not straight line.

So if we [assume we were investing adult then] were to go back to 1950, if you were to invest in 20 year bond [assuming there is 20 yr bond] in a rising interest rate environment for the next 31 years. For performance basis comparison made between [1] Stocks, [2] Bonds, [3] GICs, [4] Inflation. Let me first talked about inflation, if we are comparing all the 4 comparing investing classes assuming inflation is an investing class for comparison purpose. If all the investing asset class were to have initial $100 investment and reinvest the 'yield' returns similar to DRIP used in stocks. For comparison for the next 31 years along the way taking 5, 10, 15, 20, 25, 30 yr, readings of MARKET VALUES of each asset class; the actual case in history [from memory] is that bonds mostly can not even beat inflation because of capital losses along the way and GICs performed better because there is no capital loss at any point in market value of holdings and also beating inflation. Stocks as investment class was a runaway success story.

Of course, in a falling interest rate environment, bonds will most likely be a runaway success story because of capital gains plus starting with a very high bond yield. But to start from a very low interest rate environment, for rising rate over a prolonged period:

1. Ask yourself, do you have the stomach to see that your bond investment is underwater compare to inflation?
2. There may also be some point in time that you stock holdings go for a tail spin when you bond holdings are also in the dog house.


For the historic comparison as in Andex Chart, long bond is used for comparison. Even if one were to use ladder bonds, in a prolonged period of rising rates environment, the market value at any point in time will still be below that of GICs. But when long bond is used, because of the compounding effect, the stocks outperformed bonds by a HUGE margin in dollar value.

[Shakespeare]

the market value at any point in time will still be below that of GICs.

There are two effects that must be considered:

1. The absence of 'mark-to-market' on GICs. This effect is an illusion.

2. A shorter duration of a GIC ladder (~2.5 years) than a broad-based bond fund or index (which has a longer duration, depending on rates).

If the durations are similar, the returns will be similar.

[Justise]

[If the durations are similar, the returns will be similar./quote]

1. The absence of 'mark-to-market' on GICs. This effect is an illusion.

If you were to have leverage with margin, with a equal weighting on bonds, GICs, and stocks, your holdings will be marked to market and if both bonds and stocks are underwater badly, your banker will still mark-to-market for GICs as having no drop in value. Given a choice, obviously you would prefer to relinquish GIC for no comfirm loss while allow the others' paper losses a chance to recover.

Even if the the durations are the same, and the yields are the same, you have no trading costs for GICs, whereas there is trading costs on top spreads between bid and ask for bonds. If taking that into account, over the long term, bonds will definitely trail GICs. Even the Andex Chart does not taking bond commissions into account, in a rising rate environment, bonds at best just tango along with inflation.

[queerasmoi]

If the math is so simple, let's pull those historical numbers and see them. I don't know where to get those numbers myself. Let's say, for all the 5- and 10-year windows ranging from 1950-1990, what was the total return on a diversified bond portfolio of a 5-year duration? Perhaps they can be plotted as a histogram so we see how good/bad some of those windows were. (bump the end date up to 1990 because we have to factor in the fact that some time windows in what we *think* are rising-rate periods will actually have dropping rates too).

I'm too busy getting my Chemistry degree to engage in such antics...

If you don't have the stomach to see your bond holdings fail to beat inflation, you could go for RRBs. You may not beat inflation by much at today's real yields, but at least you'll have some certainty.

[scomac]

If the math is so simple, let's pull those historical numbers and see them. I don't know where to get those numbers myself.

Justise is referring to numbers that he has derived off of an Andex Chart. Here is a link describing what he's talking about. Sample 2010 Andex Chart.

[Added] From the chart you can see that the total return of the 5 year GIC (beige line) outpaced the DEX long bond (gold line) during the period of rising interest rates from 1950-80. The problem with the argument is that it is an on-going comparison of a 5 year term versus a 30 year term which does not reflect the reality of a typical fixed income portfolio which has a range of maturities that are constantly maturing and being reinvested at the going rate. The only way to duplicate the return of the chart would be to buy a 5 yr. and a 30 yr. bond and then turn them over each year to maintain the term length. Thus the logic that short term debt will outperform long term debt in a rising rate environment is flawed as the snapshot being used as evidence is not reflective of real life fixed income portfolios.

[Shakespeare]

I believe Bruce did a study on GIC returns some years ago and found they 'won' over a certain period.

[queerasmoi]

Interesting chart, thanks. But yeah I'd like to see numbers that are reflective of actual portfolios.

The behaviour of the one bond index shown there is particularly extreme seeing as it's a long bond index. But even then, you see that every spike in interest rates produces a drop or stagnation in the bond portfolio for 1-3 years and then it starts to swing back up due to the higher yields.

[Justise]

Interesting chart, thanks. But yeah I'd like to see numbers that are reflective of actual portfolios.

The behaviour of the one bond index shown there is particularly extreme seeing as it's a long bond index. But even then, you see that every spike in interest rates produces a drop or stagnation in the bond portfolio for 1-3 years and then it starts to swing back up due to the higher yields.

I think long bond index is used because long bonds in most cases are having higher yields than short bonds and therefore supposedly having a better total return when the interest rate neither goes up nor goes down or remains flat over an extended period and the coupon payments reinvested. As no ones knows the interest rates are going up or down, therefore the use of long bond is has supposedly a positive advantage over short bond. Combining with index which reduces trading costs, I think it cannot be said of "particular extreme". What matters is therefore the interest rate direction.

Regarding your quote - "you see that every spike in interest rates produces a drop or stagnation in the bond portfolio for 1-3 years" - , don't forget that what you see the drop is on LOG scale which means that the drop is very drastic on linear scale for 1 - 3 years [for some 6 years in one case] and the spike in interest rate is usually because of spike in interest rate. And therefore, inflation curve is compounded at accelerated rate i.e. the inflation potfolio raced ahead while the bond potfolio dropped. As you said, those are actual potfolios.

[queerasmoi]

Well, when I say actual portfolios, I mean that I have rarely heard of an investor placing *only* long bonds in their asset allocation. Most FWFers who use a bond fund will either use a broad index including a variety of durations (such as XBB, avg duration around 6 years) or a short bond index (such as XSB, avg duration around 2.5). Those FWFers who hold bonds directly either have a 5- or 10-year bond ladder, or are holding long bonds expressly because they want the income stream and will not sell them before maturity. Presently the long bond index (XLB) has a duration around 13 years.

So ultimately what I'm getting at is, I look at that chart, and I don't "bubble-like" behaviour in bonds at any point in time. Hold the long bond index for any period of several years and you will get returns that are consistent with fixed-income. Hold a *shorter* bond index or a bond ladder, and the portfolio will recover faster but maybe have slightly reduced returns.

Your original question was, should some of your bond portion be reallocated to equities, based on a belief that bonds will drop? I still think the answer is no - they don't serve the same purpose in a portfolio. The split equity/bond portfolios on that chart seem to do just fine after periods where bonds have dropped; the rebalancing probably helps out.

If you want to ramp up your equities then you have to ask if it would be appropriate for you and your risk tolerance. You have no way of knowing whether interest rates will rise incredibly rapidly in an upwards equity market (1978-1981), or maybe somewhat less rapidly in a falling equity market (1972-1974), or very slowly in a stagnant equity market (1987-1990) and so forth. Equities are still risky, and we can still expect them to be more volatile than bonds. If you can block out all the noise about what you and everyone else thinks the market will do, and still conclude that you can stomach more equities, then do it. And good luck.

[Taggart]

I believe Bruce did a study on GIC returns some years ago and found they 'won' over a certain period.

My apologies, but I could only find a free brief abstract to the excellent column from James Daw and Bruce Cohen that Shakes is obviously referring to. GIC investing measures up.

[CROCKD]

I am wrestling with where I should be headed with my asset allocation. Attempting to diversify my RRSP 3 to 4 years ago I bought XIN and a bank European fund. These are both down about 30% and I had intended to sell them last year but everything went on hold as a result of my serious illness. They represent about 6% of my RIF.

I turned 71 in 2010 and must begin making substantial withdrawals from my RIF.(7.38% minimum this year). Fortunately I don't anticipate needing the income to live on, so have an option of making withdrawals in kind. However my income requirements will be higher as I must find the money to pay the tax bill and especially as I have estimated that a portion of my OAS wll be clawed back.

My RIF assets consist of 56% equities and the remainder mostly fixed income in the form of bond ETF's.
I also have a LIF containing bonds and bond ETF's worth approximately 1/2 of the RIF.

I have a substantial position in cash in my non registered assets ( about 2/3 of value of RIF/LIF assets) with the returns that that currently implies.

Realizing that everyone is different, does anyone have an opinion of what I should be doing with asset allocation in my portfolio at this stage of my life.

[like_to_retire]

I can't advise on your allocation, but it seems like a good idea to transfer out in-kind the depressed shares of XIN and the European fund to satisfy the RRIF withdrawal requirement while those shares still hold a low value. It will take more of those shares to do so than it would if they were at purchased cost base or higher. Then when those shares hopefully appreciate outside the RRIF, you will have the result of better tax efficiency for those shares.

ltr

[CROCKD]

Thanks for your suggestion ltr.
I don't usually make New Year resolutions. But one of them is to appreciate the efforts of others on my behalf.

Indeed I think I will use your suggested philosophy for my first year withdrawal.

[Justise]

I have a strong supporter of my view on bonds with the headline news - Bonds among most dangerous assets: Warren Buffett

[Justise]

Please read this website article - http://ca.finance.yahoo.com/news/why-bo ... 50900.html

The authors explained in layman language why bonds are riskiest asset class.

[flywaysuzy]

Well, I read and am perplexed as to why the teeter totter metaphor was chosen... he says bonds are way overpriced. Recently we've seen Italian and Greek bonds being sold with fairly hefty interest rates attached, but nobody seeking safe investments are going to be buying these, only gamblers or people prepared to take a big hit on the face value of these bonds are going to be purchasing them. I think bonds in Canada are much lower, although I haven't checked in a while. If they had a decent return I would have bought some!

He says investors should invest their money in something that is insulated and provides a good return but neglects to mention what this miracle investment might be...

Maybe there's a sequel coming up that will explain further...

[j831robert]

Sensie: Methinks your middle para in your offering of 7 March above "I think net worth has a decided influence ......" is a gem (but perhaps you might have left the statement at that). In this day and age perhaps a shotgun and good supply of ammunition for it might be the best investment for survival against a total collapse of our economy.