Fixed income ladder

A fixed income ladder is created by dividing a sum of money into several portions (usually equal) and investing them in bonds or guaranteed investment certificates (GICs) of varying maturities (usually equally-spaced). For example, a "Five-year GIC" ladder" could be constructed by dividing $50000 into five $10000 portions and using the portions to purchase GICs with maturities of one, two, three, four, and five years.

GIC ladders
GIC ladders are composed entirely of GICs. Since a federally-insured GIC is limited to a maximum term of five years, all federally-insured GIC ladders must have a maximum term of five years or less. However, longer maturities can be accommodated if Guaranteed Interest Accounts (GIAs) are included, or if other types of deposit insurance are acceptable.

Investors with large portfolios may wish to take care that the selected GICs are within the maximum limits of the deposit insurance used. Note that GIAs are guaranteed by the insurance organization Assuris, not the Government of Canada.

As a study by Mawani, Milevsky, and Landzberg showed, investors in high tax brackets who want significant real, after-tax returns should not use a GIC ladder in a taxable account.

Bond ladders
Bond ladders are similar to GIC ladders, but are composed of government or corporate bonds. Since bonds can have maturities of up to 30 years, bond ladders can be (and usually are) designed with a longer term than GIC ladders. Laddering has been recommended by Cunningham. However Hymas feels that only those with large bond portfolios (one million dollars or more) should invest in bonds directly.

Since the underlying bonds are not guaranteed (unless they are restricted to Government of Canada bonds), more care generally needs to be taken in the selection of the ladder rungs than for GIC ladders.

Rolling and declining ladders
Fixed-income ladders may be rolling or declining.
 * A rolling ladder reinvests each maturing bond or GIC at the longest maturity; for the five-year GIC ladder above, a new five-year GIC would be purchased once a year. This "rolling over" slowly adjusts the ladder for interest rate changes.
 * A declining ladder modifies the new bond or GIC purchases so that they mature at a fixed maturity date some time in the future, until the entire ladder matures at a given future date.

Rolling ladders are used to remove interest rate sensitivity from the investment decision. Declining ladders are used to fund specific needs (such as education costs) at the date they mature.

Designing a ladder
The first step in designing a ladder is determining whether it is to be rolling or declining. Once that decision is made, it is then necessary to decide whether interest is to be compounded or withdrawn. For a compounding GIC ladder, compounding GICs will be used. For a compounding bond ladder, strip bonds can be suitable if the ladder is in a registered account. For non-compounding ladders, non-compounding GICs or conventional bonds would be used.

Example GIC ladder
To establish a five-year GIC ladder with a value of $50000, the first step after the type of GIC has been established (this example will use annual-rate GICs) is to consult the GIC rates offered by the investor's brokerage. For example, a GIC ladder constructed with five $10000 rungs each with the best annual rates offered by BMO InvestorLine on December 14, 2012 would have the following structure:


 * {| class="wikitable"

! GIC Term ! Amount ! Rate
 * 1 year||$10,000||1.70%
 * 2 years||$10,000||2.15%
 * 3 years||$10,000||2.20%
 * 4 years||$10,000||2.25%
 * 5 years||$10,000||2.40%
 * ||Average:||2.14%
 * }
 * 4 years||$10,000||2.25%
 * 5 years||$10,000||2.40%
 * ||Average:||2.14%
 * }
 * ||Average:||2.14%
 * }

Since the ladder rungs are equal in size and spacing, the simple average of the term and yield can be used. If the amounts or spacing were not equal, a weighted mean (with the weights in accordance with the dollar amounts) would be required.

Using the average coupon of 2.14%, an average maturity of 3 years, and annual payment, the duration of this ladder can be estimated as 2.94 years upon creation. However, it will drop to about 2 years just before the maturing GIC is rolled over. Since each rung is renewed at the maximum term when it matures, in five years the entire ladder will yield the five-year rate - in this case, 2.40%, if interest rates remain unchanged.

Similar calculations can be made for bond ladders.

Comparison with a bond exchange-traded fund
Once the average yield, average term, and duration of the ladder have been calculated, it is possible to compare them with similar bond exchange-traded funds (ETFs) or mutual funds. Since a short-term ladder has been constructed, it should be compared to a short-term bond ETF. Using the iShares Canadian Short Term Bond Index ETF (tsx: XSB) as an example, the following characteristics can be obtained:
 * Weighted Average Yield to Maturity (YTM) of 1.54%
 * Weighted Average Duration of 2.69 years
 * Management Expense Ratio (MER) of 0.28%

A common method of estimating the expected return of the ETF is to subtract the MER from the YTM. However, the index is constructed in such a way that it can take advantage of a technique known as "rolling down the yield curve" by selling bonds when the maturity reaches 1 year. The effects of this change will depend upon the shape of the yield curve, but will usually increase the ETF's return. The amount of increase at current yield levels has been estimated as 0.5% but depends on the ETF composition, the yield curve, and yield curve changes, and can not be known in advance. Nonetheless, the purchaser of this bond ETF would give up a significant fraction of his or her prospective yield compared to the certainty of the GIC ladder in order to obtain greater liquidity.

Similar comparisons can be made with other bond ETFs or bond funds. Note that the comparisons should be between investments of similar average duration and credit quality so that the effects of interest rates are similar.

Credit quality and default risk
A GIC ladder has rungs which have a government guarantee, and can fairly be compared to a ladder of Government of Canada bonds, in comparison to which it will generally offer significantly higher rates. A bond ladder, however, can contain higher-yielding but lower-quality bonds. Therefore, bond ladders can differ in credit quality and default risk. In general, any comparison between a bond ladder and a bond fund or ETF should take into account differences in the quality of the holdings. The default rates have been estimated by Hamilton and Ou.

Effect of interest rates
Bond ladders, like bond ETFs and bond mutual funds, are aggregates of individual bonds and are subject to the same pricing effects with interest rate changes. As with individual bonds, there is an inverse relationship between pricing and interest rates: a rise in interest rates causes a drop in pricing and vice-versa. This is addressed in more detail in conventional bonds: inverse relationship.

Once the duration of the bond ladder, ETF, or fund is known, the pricing change of the aggregate can be estimated. In the above example for XSB, the duration of 2.69 years means that a 1% rise in interest rates would cause a price drop of about 2.69%.

Pricing and reinvestment risk
A common misconception on the comparison between fixed-income ladders and bond funds or ETFs arises from the price drops that funds or ETFs can incur should interest rates rise. The ladder investor claims that, since he or she got "100 cents on the dollar" from a maturing bond or GIC, the ladder is immune to such effects.

Unfortunately, this misconception ignores the effects of accumulated interest and reinvestment risk. Suppose that interest rates have gone up. That means that the accumulated interest from a maturing bond or GIC would be less than the current rate. Thus, less total cash is available for reinvestment - which, for a rolling ladder, must be at the maximum duration and the highest interest rate sensitivity.

Reinvestment risk in a ladder thus largely cancels out the pricing changes in a bond fund or ETF. A more detailed analysis of the similarity has been made by Hymas.

GICs and market pricing
Since there is no active secondary market for GICs (which, unless cashable, can generally only be sold with difficulty and at a discount), brokerage statements usually show the nominal GIC value as the sum of face value and accrued interest. This leads to the common but mistaken perception that a GIC ladder is unaffected by interest rates. In fact, if there were a liquid secondary market, the market prices of GICs would react similarly to those of bonds with similar fundamentals. Nonetheless, the pricing of a GIC ladder is readjusted when the maturing GIC (which has the lowest interest rate sensitivity) is used to purchase the longest-maturity component of the ladder, which has the highest interest rate sensitivity. This partially corrects for the lack of market pricing.

It follows, from these arguments, that fixed income aggregates - whether ladders, funds or ETFs - react similarly to interest rate changes, although that reaction may not be visible to the investor.

Disadvantages of ladders
Although in certain cases (particularly the GIC ladder), fixed income ladders offer a yield advantage compared to similar funds or ETFs, they also have some disadvantages.


 * The ladder is significantly less liquid. For a five-year rolling GIC ladder with non-cashable GICs, only 20% of the funds can be available, and then on only on one day each year if funds are reinvested immediately. A GIC ladder is thus not suitable for emergency funds.


 * The ladder generally will have far fewer holdings than a bond fund or ETF, and thus has much higher risk granularity. Although institutional failure is covered by deposit insurance with a GIC ladder, there is no such coverage with non-government bonds. Thus, should a bond default occur, a significant portion of the bond ladder could be put at risk.


 * Although the ladder holder avoids the MER of a fund or ETF, he or she may pay significantly higher prices for bond purchase, negating some of that MER advantage.

Performance
In a US study, Barnes et al. (1984) examined the performance of various corporate bond management strategies between 1969 and 1981. Although yields rose dramatically between 1979 and 1981, the study period contains 14 quarters of falling interest rates. Thus the strategies can be evaluated during periods of both rising rates, and falling rates. Strategies studied include equal allocation into all available issues, with quarterly rebalancing; long bonds only; short bonds only; a barbell; a 15 year ladder; a buy-and-hold strategy; a 'Fisher and Weil duration' approach; and investing in bonds with low credit ratings. Taking the whole period, the laddered strategy performed among the best (although the differences in performance are not statistically significant). When rates were rising, it was the third best strategy. And when rates were falling, laddering was in the middle of the pack. In summary, laddering performs reasonably well when compared to other strategies.