Canadian Money Forum banner

Running my own bond fund, comparing to VAB

92K views 195 replies 25 participants last post by  Covariance 
#1 · (Edited)
Starting September, I began using a dedicated account for my fixed income. I'm trying to see if I can replicate what VAB does, but with direct ownership of bonds. My portfolio is similar, around 10 year maturity with mostly govt bonds plus a few GICs.

I'm comparing to my broker-generated monthly performance report, which includes fees and isn't affected by cashflows. Current results seem to be tracking VAB well, but it will take at least a year of data to know.
 
#79 · (Edited)
There are those who know a lot about bonds in this board, likes j4b. However, many are learning, like myself, so the following might be useful.

"The Sharpe ratio has been about 0.40 at the one-year maturity but falls to about 0.26 if we extend the maturity to about five years, and it continues to fall as we extend the maturity. Thus holding assets with a maturity of about one to two years is the prudent strategy for those investors wishing to maximize the risk-reward relationship."

Swedroe, Larry E.; Hempen, Joseph H.. The Only Guide to a Winning Bond Strategy You'll Ever Need: The Way Smart Money Preserves Wealth Today (pp. 68-69). St. Martin's Press. Kindle Edition.

Larry Swedroe doesn't give more detail, but the context of the quote would suggest that the data from which the analysis is derived is historical US federal government bonds. Of course, Sharpe ratios are not an ideal measure of risk, but they are worth looking at.

It makes me consider rolling over 1 year GICs or 2 year GICs.

http://www.financialpost.com/personal-finance/rates/gic-annual.html

From the above link, the best rates for 1 year, 2 year and 5 year GICs are 2.1%, 2.050% and 2.5% respectively.

So a strategy of rolling over 1 or 2 year GICs might be a reasonable option. It would have less reinvestment risk than a HISA. It would have less liquidity risk and less inflation risk than a 5 year GIC ladder. And at least at present, you're not sacrificing much yield by having shorter term exposure.
 
#82 ·
With the remark "And at least at present", one could accuse me of fixed income market timing. But I doubt that my conclusions would change regardless of the yield curve. Although Larry Swedroe doesn't mention the time period on which the data is drawn, my guess is that it's a long one. And the strategy of rolling 1 or 2 year GICs is for money needed within the next 3-5 years. I have earlier talked about using a HISA for money needed within the next 3-5 years. A strategy is rolling 1 or 2year GICs is just going a little further out on the yield curve.
 
#80 · (Edited)
^ I understand (and somewhat agree) with all these arguments about targeting the short end of the yield curve.... right now. No question, as of Wednesday February 15, the best risk/reward tradeoff is in the short end.

My question though, is when do you adjust to farther down the yield curve? Everyone on these forums is saying basically the same thing: "current conditions make the short end of the curve the best fixed income investment. Some time down the road I will go back to the long end of the curve when it becomes more attractive".

OK, so hum along in cash and the shortest term fixed income, forfeiting performance as you do so. Luckily you timed it so well, bond yields immediately go up, and then Mr. Market rings a bell and tells you that bonds have stopped falling and now you can go far down the yield curve -- really? Does that sound plausible over the span of several decades?

This is bond market timing. It doesn't sound like market timing to people, for some reason. People have been saying this for years, at least back to 2009 when yields first fell quite low. For the last 8 years, people have unsuccessfully timed the bond market. That's a pretty long streak of timing failure.

Here for illustration is the chronic failure to "out-smart" the bond market. Those short term bonds XSB (blue) that were so obviously the best investment, according to everyone, have underperformed XBB (green). Which is exactly what you would expect if you have a neutral rate outlook, or even if you allow for mild rate increases: http://stockcharts.com/h-sc/ui?s=XBB.TO&p=D&st=2009-01-01&en=(today)&id=p76152527729

Again -- why do we all recognize that we can't time the stock market, but we think we can time the bond market? Not only are you timing the bond market, but you're putting a lot of faith in a very specific scenario: rapidly rising interest rates and an end to a 36 year interest rate trend.

You might have better results than me, but I don't believe that I can time the bond market. This is why I stick to a certain reasonable long horizon maturity (VAB and its 10 year average) or the maximum CDIC insured GIC term (5 years).

The simple reason: with a neutral rate outlook, or even if you allow for mild rate increases, VAB/XBB (10 yr maturity) will outperform cash/GIC/XSB (short maturity) over decade-ish periods.

It's the couch potato strategy. You assume that you can't time or outsmart the market, and decide to just take exposure to the asset class.

Even if you are correct that interest rates are about to go higher, that's not enough. You also have to successfully time your entry back into regular (VAB) bonds. Get this timing wrong (e.g. everyone over the last 8 years!) and you dramatically underperform.
 
#81 ·
I think that there is a 69 year span in the USA in the 19th century where bonds outperformed stocks. That's something to think about.

This is from p.95 of "Stocks for the Long Run". This shows highest and lowest real returns on stocks, bonds and bills over 1, 2 5,10, 20 and 30 years holding periods in the USA between 1802-2012.

Worst 1 year return for stocks is -38.6%, for bonds -21.9%. 2 years -31.7%, -15.9%, 5 years -11.9% -10.1%, 10 years -4.1% -5.4%, 20 years 1.0% -3.1% 30 years 2.8% -2.0%

By 5 years, the worst real returns are similar for stocks and bonds.

The following is from p.95 of "Stocks for the Long Run". This shows highest and lowest real returns on stocks, bonds and bills over 1, 2 5,10, 20 and 30 years holding periods in the USA between 1802-2012.

Worst 1 year return for stocks is -38.6%, for cash -15.6%. 2 years -31.7% -15.1% , 5 years -11.9% -8.3%, 10 years -4.1% -6.1% , 20 years 1.0% -3.0% , 30 years 2.8% -1.6%

If your time horizon is 1-5 years, cash and bonds are safer than stocks. But if you have a time horizon of 5-10 years or more, stocks are safer than cash or bonds.

That's data from 1802-2012. So one can't be accused of choosing time periods favorable to your conclusion. In fact, the data is biased in favor of bonds. The older data includes the gold standard time period. During that time, deflation was much more severe than it has been in the paper currency era. Government bonds will tend to do better than stocks, in the face of deflation.

Finally, the above returns ignore taxes. In a taxable account with a 50% marginal tax rate, bonds have a handicap that is difficult to over come.
 
#83 ·
^ Interesting, but:

1. Not a huge discovery that long-term bonds can be just as volatile as stocks. I don't know what the above statistics actually covers, but the 19th century data are long-term bonds. Assume that TIPS and other inflation-linked instruments are not covered at all.

2. Not sure we can claim the data are biased in favour of bonds. The 20th century was also special; the loss of gold standard and the time it took governments to figure out inflation resulted in major damage to bonds. There is nothing to indicate that 20th century conditions will be repeated.

All this history is very interesting but in truth... We just can't project the past forward. Diversification is good in the game of "winning by not losing".

Having said all this, I just don't see how the risk is priced within any bonds on offer right now and the delta between government bonds and private/junk bonds does not seem to be anywhere near historic standards. It's probably because I don't understand the instrument but I tend to be cautious about the things I don't understand. The pricing is moving in the right direction though.
 
#86 ·
"Decide on your asset allocation and stick with it."

This is something I just don't see. If you were running a business, you wouldn't choose a mix of products to sell and stick with it regardless. If one product was going out of favour, you would use your floor space to display a product that is likely to sell.

Economies change and it seems to me that we as investors must change too. Early in past century, railroads represented over 60% of the US market. Now only about 0.2%. In 1900 IT did not exist and now it is about 23% of US economy. Banks & Finance share of US markets has doubled.

So I could see deciding on your asset allocation and sticking with it for at most 12 months. Then reassess.
 
#88 · (Edited)
A reference to back up my claim that persistent allocations are a better strategy than "tactical asset allocation"
http://www.etf.com/sections/index-i...are-tactical-asset-allocation.html?nopaging=1

I think that some people in this thread are basically proposing tactical asset allocation, where they are looking at current bond market yields and inflation rates and determining that bonds should be cut out of the portfolio because they are projecting poor returns ahead. As good as this kind of idea sounds in theory, it doesn't work well in practice.
 
#89 ·
Don't believe there is any serious source recommending that there should be no fixed income at all, but prominent proponents of passive such as William Bernstein do recommend to time your bond purchases. For example in the Four Pillars he is recommending only purchasing government bonds during times like today.
 
#96 ·
I'm not suggesting anything. Maybe there are real humans behind your usernames, but for all I know I'm talking to a bunch of 1's and 0's in the matrix. I wouldn't give financial advice to someone I know nothing about. Just explaining to j4b why I personally don't see value in bonds at the present.
 
#100 ·
James,
I don't disagree that is hard to get market timing or allocation timing "right". For those who are working or heavily involved with other things, some sort of autopilot investing using some of those often espoused "rules" may be the right thing. But if you have time and do it in a way, that limits or avoids chance of a loss if you are wrong, then I see nothing wrong with trying. I am sure most managed funds and professional investors attempt to get timing/allocation right.
 
#101 ·
One anecdote from me: I knew that stocks were overvalued in 2006-2007 and there were financial sector risks, so I (correctly) lightened up on stocks. This was market timing at its finest ... I avoided the 2008 catastrophe.

BUT after the turmoil, I failed to buy back in. Instead if I had been steadily invested in SPY & XIU, that would have done much better.

I similarly have tried to outsmart the bond market. Even in the early 2000s, I was limiting myself to 3 yr instruments thinking that interest rates would rise. I was wrong. Just like with my stock experience, I would have done much better if I just stayed consistently invested in the asset class.
 
#102 · (Edited)
https://www.youtube.com/watch?v=_chiIIxMGl0&list=PL24qYBiXaDDsSGgzBFxEe8SsECXiqVFpI

Above is a link to a video by Lars Kroijer. Basically, it's about indexing, which isn't relevant to this thread. But what is relevant is his approach to asset allocation. He advocates what some might call a barbell approach. The equity portion of the portfolio is indexed. But the rest of the portfolio is in cash or government bonds. It's his advocacy of cash for the nonequity portion of the portfolio that I find interesting.

http://www.financialpost.com/personal-finance/rates/gic-annual.html
http://www.bankofcanada.ca/rates/interest-rates/canadian-bonds/
http://www.financialpost.com/markets/data/bonds-canadian.html
https://www.highinterestsavings.ca/chart/

From p. 95 of "Stocks For The Long Run", the following are the worst real returns from 1802-2012 for US bonds and bill over 1 ,2 and 5 year holding periods respectively: -21.9% vs. -15.6%, -15.9% vs -15.1% and -10.1% vs. -8.3%. What defines bonds is important, because a 5 year bond behaves quite differently than a 30 year bond. Nevertheless, if capital preservation is important, bills look like a reasonable option.

The following is only about products with CDIC backing. The top HISA is 2%. The top 1 year, 2 year and 5 year GICs are 1.75%, 1.95% and 2.50% respectively.

1 month Canadian treasury bills, 1 year Canadian Treasury bills, 2 year government of Canada bonds and 5 year government of Canada bonds are paying 0.44%, 0.60%, 0.79% and 1.18% respectively.

A province of Ontario bond maturing on March 8/2018 has a yield of 0.78%. Province of Quebec bonds maturing on December 1/17 and December 1/18 have yields of 0.70% and 1.00% respectively.

Enbridge Pipelines has a bond maturing on Nov 19/18 with a yield of 1.37%. Rogers Communications has a bond maturing on March 22/21 with a yield of 2.06%.

As mentioned previously, I see the role of cash/fixed income as a way to offset known and unknown liabilities for the next 3-5 years. A bond/GIC approach is a good way to meet that goal. But I'm becoming increasingly convinced that a HISA is a reasonable alternative to a bond/GIC strategy.

What strikes me is how simple it is. To become a fixed income expert takes a lot of time, and even if I wanted to spend the time, I'm not certain that I would succeed. But I think I can become an expert about Canadian HISAs, and it won't take that much time. And I doubt that I will significantly underperform strategies that are much more sophisticated.
 
#103 ·
But I think I can become an expert about Canadian HISAs, and it won't take that much time. And I doubt that I will significantly underperform strategies that are much more sophisticated.
Let's look at a period where interest rates have actually gone up. On 2012-07-24 the govt 10yr was 1.58% and today it's 1.74%.

According to your thinking, the period from 2012-07-24 to 2017-02-17 should be a period where cash/HISA outperforms bonds. XBB annualized performance in this period was 2.51%.

Even at the sketchiest mortgage trust high interest savings accounts you could have found over that period, I don't think you would have outperformed in HISAs. Some of the very high rates we see today are promotional and from new subsidiaries of places that didn't even exist before. A place like Outlook Financial has been a long established top HISA yield. Over this period the average Outlook Financial HISA rate was 1.85%

So here is a period where interest rates went up -- just as you are fearing -- and XBB/VAB returned 2.51% vs about 1.85% in HISA. Personally I would be happier with the 2.51% from a static "set it and forget it" investment, vs constantly shuffling money between different sketchy mortgage lenders, opening endless new accounts, and catching promo rates.

I don't see how HISAs are less work, and they actually underperformed during this period where interest rates went up!
 
#104 · (Edited)
You might say, well the 1.85% in HISA during that period wasn't too much worse than 2.51% in XBB -- and I agree, they both round to 2 more or less.

As far as predicting what's in store for the future,

1. bond yields go up gently -> VAB outperforms
2. bond yields go up sharply -> HISA outperforms
3. bond yields go down -> VAB outperforms
4. bond yields stay flat -> VAB outperforms

Personally when I look at that matrix of forecasts, I say to myself: the average case is that VAB outperforms. And holding VAB or XBB is much easier than jumping around between HISAs to get the best rates.
 
#105 ·
The purpose of a HISA is to offset cash flow liabilities over the next 5 years. XBB has an average effective duration of 7.16 years; 10.75% of the fund is in BBB bonds. I wouldn't use XBB to offset cash flow liabilities over the next 5 years. With XBB, you're taking on interest rate risk and some credit risk. You should outperform a HISA. I'd prefer to take my risk in equities, rather than bonds, where it historically has been better rewarded. And that's ignoring posttax returns.

http://quote.morningstar.ca/QuickTakes/ETF/etf_Portfolionew.aspx?t=XBB&region=CAN&culture=en-CA
 
#108 ·
My portfolio's weighted average maturity/term is 7 years. This is shorter than VAB at 10 years and explains why my portfolio of individual bonds is less volatile than VAB.

This has been good for me during this soft period in bonds, but it also means that I will underperform VAB when bonds strengthen (e.g. February).

Currently I am debating whether to immediately boost my maturity upward to match VAB, or to attempt some market timing [*] and keep my avg term a bit lower until bonds weaken more.



[*] I doubt that I can time the bond market, every though most CMF participants seem to be sure they can time bonds
 
#110 ·
Fund duration risk

My portfolio's weighted average maturity/term is 7 years. This is shorter than VAB at 10 years and explains why my portfolio of individual bonds is less volatile than VAB.
Great thread, and your work at running your own private bond fund is inspiring James! Unfortunate that rates are so low and we face the risk of rising rates. I've yet to complete reading the entire thread, but I'm curious about duration risk and the impact of rising rates ahead.

With a fund like your fund, ZAG, VAB, or XBB ... a growing concern is duration risk that the fund value will drop as rates rise. ZAG has a duration of 7 similar to yours. The US Fed is expected to raise rates about 3 or 4 times during 2017 for approximately an increase of .75 - 1%. Even though its US rates, it obviously effects rates in Canada when you study the charts.


example: www . tradingview . com/x/B6jky3uL/

With the frist 1/4% rate hike in Dec 2015 the ZAG pulled back 2.6% and recovered. But with a series of 1/4% hikes in the Fed funds rate approximately every three months from Dec 2016, should these 7 year duration funds not be expected to drop about 7% in fund value by the end of the year?
 
#109 ·
I found a better way to track my portfolio's performance vs VAB. This chart is built using monthly performance from both Vanguard and my brokerage; total returns. Curious to see if I can keep up with the professionals at Vanguard.

After six months I'm tracking VAB perfectly so far

Line Slope Plot Parallel
 
#113 ·
Another example. Between Jan 1994 and May 1995, the Fed increased rates +3.0%. The 10 year treasury bond yield increased +1.3% ... http://stockcharts.com/h-sc/ui?s=$TNX&p=D&st=1994-01-01&en=1995-05-01&id=p35931391770

Bond funds similar to VAB saw a temporary 6% decline but ended positive 3%, as shown here ... http://stockcharts.com/h-sc/ui?s=VBMFX&p=D&st=1994-01-01&en=1995-05-01&id=p94689011619

Notice that the temporary losses in bond funds were nowhere near as severe as someone may have predicted by saying 3.0 rate hike x duration of 6.0 = 18% decline. And after this 15 month series of Fed rate hikes, the bond fund was ultimately higher by 3%
 
#115 · (Edited)
http://www.sensibleinvesting.tv/doc...8efe-79ee5a4055f2/acuity issue16 sensible.pdf

The above link is from the website sensibleinvesting.tv. For a novice investor, it's a good website.

The following isn't directly relevant to this thread, but it is to novice bond investors.

UK BBB bonds, from Nov 2007 to Feb 2009, declined 15% in value. I knew that junk bonds did poorly at that time. And I knew that high quality bonds increased in value during that time. But I didn't know what happened to bonds in between the two categories. BBB bonds are considered investment grade. It would be interesting to see what happened to A bonds during this time period. If BBB bonds were down 15%, my guess is that A bond might very well have had a modest loss during this time. If you're using bonds to diversify equity risk, then perhaps the minimum credit quality should be AA.

Edited to include the following. During the same time period, UK high yield went down 24% and UK stocks went down 41%. Were BBB bonds greatly different than junk bonds?
 
#116 ·
http://www.servowealth.com/blog/be-careful-with-your-bonds

This is a followup on my last post. It seems that it's been too long to edit my last post, so I have to start a new post.

This is about what bonds went up in 2008 and what went down.

The link above shows US intermediate bond returns in 2008. High yield was down around -26%. Baa around -8%. A around -5%. Aa around +2%. Aaa around +8%. US Government (I assume Treasuries) around +10%.

There's a clear message here. In a flight to quality, the bonds that will diversify stocks are likely to be bonds with credit ratings of at least Aa. And you can make a case that Aa bonds didn't diversify stocks in 2008, although they did dilute them, just as cash would have done.
 
#119 ·
This thread was intended for me to track my bond portfolio, but you guys keep dragging me into bond debates.

Long term, Canadian bonds have shown solid real returns and certainly provide a role in portfolio diversification even at times (like the current instant) where their real returns are low.

You cannot predict bond prices. The consensus around here is to "avoid bonds and wait until rates are better" -- which is just a market timing attempt.

A bond fund, as opposed to an individual bond, provides ongoing exposure to bonds and the returns they generate. If interest rates go up, you benefit from an increase in returns over the long time. Many people misunderstand what a bond fund is, and confuse it with the issue of an individual bond, which certainly does plummet when rates rise. The key difference is that a bond fund maintains a constant maturity.
 
#121 ·
James,

I hope this isn't considered another diversion of the topic of this thread.

I was wondering what software (or otherwise methods) you use to keep track of your custom bond "ETF" and more importantly how do you compare it to VAB?
I am running my own REIT basket and wanted to use ZRE as a benchmark so I thought you might share some of your experience in this regard. In particular how do you compare beta, volatility, etc.

Cheers,
JC
 
#122 ·
Hi John, the iTrade performance calculator gives me the monthly % change. I then just use a spreadsheet where I list my monthly % gains vs the Vanguard monthly % gains, which you can find in the first posts of my thread.

The most recent graphs I posted are just the cumulative gains by calculating, in a spreadsheet, the total return of the cumulative months together
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top