Road to ruin

While they’ve been around for just over two years, Guaranteed Minimum Withdrawal Benefit (GMWB) segregated funds have made a significant impact on the Canadian marketplace. The introduction of an investment vehicle that offers the opportunity to participate in investment markets while guaranteeing a stream of income for life has been so attractive to the marketplace that it has grown from nothing to over $10 billion in its short life.

The product guarantees a client 5% of his or her deposit for life even if the account value hits zero (the ruin scenario). This amount is adjusted upwards if the portfolio value is higher than the guaranteed value on the specified reset dates (every one to three years) and downward for any withdrawals beyond the designated 5%.

However, one exception lets the investor receive more than 5% of the portfolio’s value during the course of a year – when GMWBs are owned within a RRIF account.

At issue are two competing percentages. The first is the GMWB’s 5% guarantee. The second, the government-mandated RRIF minimum withdrawal percentage. All RRIFs are, of course, subject to the government’s schedule of minimum withdrawals, which every RRIF investor must start receiving no later than age 71. This minimum payment starts off at 7.38% and climbs every year thereafter, reaching just shy of 10% at age 84 (9.93%) and a level 20% a year from age 94 onward.

So let’s consider what happens at age 71 if an investor converts an RRSP account totalling $100,000 to a RRIF and invests in GMWB funds. The RRIF schedule states he must receive 7.38% ($7,380) of the account’s value. Meanwhile the GMWB is only scheduled to pay out 5% ($5,000). Normally, this $2,380 (or 2.38%) difference would be deducted from the client’s guarantee value and would reduce the guaranteed income for the next year. This means the $100,000 deposit that the guarantee was based on would now be reduced to $97,620 and the guaranteed lifetime payment would be lowered to $4,881. This pattern would continue every year the RRIF minimum payment exceeded the guaranteed income, thereby eroding the guarantee value annually and providing little reason to pay the additional fees associated with this product. Luckily, however, this is not the case with RRIFs.

Each GMWB has written into its contract that in the case of RRIFs, the accounts will pay out the greater of the RRIF minimum payment or the guaranteed income. In the event the RRIF minimum is greater than the guaranteed amount (5%), any withdrawals above that 5% will not be considered additional withdrawals of capital and therefore won’t reduce the income guarantee. This provision creates a floor on the client’s RRIF income – in the context of the example, if the RRIF schedule stated the minimum withdrawal was $10,000, the investor would receive $10,000. If the scheduled payment was $3,000, then the investor would receive the minimum guaranteed income of $5,000 without any impact on future guarantees. Essentially this benefit lets investors withdraw more income from their GMWB investments than if non-registered funds had been used.

One thing to keep in mind, though, this does not necessarily work with locked-in RRIFs. Since these accounts also stipulate maximum withdrawal percentages, should the guarantee amount exceed the maximum locked-in RRIF payment, the lower of the two will be paid.

While the concept of having a minimum guaranteed income from a RRIF seems like a good idea, how necessary is it? In order to answer that question, we need to ask ourselves the following question: If one were to withdraw either the minimum RRIF payment or the 5% guaranteed amount, whichever is larger, what is the probability of the RRIF account ruining (running out of money)? To answer this question, I ran a Monte Carlo analysis that assumed a balanced return of 7%, standard deviation of 9% and resets every three years.

The results of this analysis are illustrated on the top left.

As we can see, based on the stated assumptions, there’s little to no chance of ruining the portfolio in the years prior to age 85. However, this quickly escalates over the next 15 years, where the probability of ruin goes from very unlikely to almost absolute certainty. These numbers would be even worse if one were to consider the possibility of a significant market downturn as was experienced last year.

The next factor to account for is mortality. After all, even if the probability of being ruined at age 100 is 95%, is insurance against that event worthwhile if so few people live that long? To find an answer, let us look at life expectancy at each of the age increments – keeping in mind life expectancy merely measures the average age expectation of a population. The worst thing an advisor could do is to plan for the client’s death at life expectancy. If they did, half their clients could run out of money. (See Tables.)

We see a 70-year-old can, on average, expect to live to age 84. At first glance, it appears there is little value to these guarantees since the probability of ruin only becomes significant quite late in life. But remember, life expectancy is an average. This means half the population of 70-year-olds will live to be older than 85.

So how many 70-year-olds will experience a significant chance of ruin in their lifetime? If we combine the findings of the tables, we can answer that question. According to the third table, almost 50% of the population will live to at least age 85 and almost one in four will live to age 90. When we combine this with the findings in the first table, we can determine:

• 50% of 70-year-olds will die prior to age 84 and experience a less than 3% probability of ruin; • 27% of 70-year-olds will die between ages 85 and 90 and experience a probability of ruin between 3% and 28%; • 16% of 70-year-olds will die between ages 90 and 95 and will experience a probability of ruin between 28% and 65%; and • 7% of 70-year-olds will die after age 95 and will experience a probability of ruin greater than 65%.

This leads to the conclusion that more than one in four 70-year-olds will experience a significant probability of their RRIF ruining during their lifetime, providing all the assumptions hold. Portfolios with lower returns, higher volatility or large market downturns will only increase the likelihood of ruining. But given the odds of ruining are no less than one in four, they are tangible and significant. And, given the consequences of losing a source of income at that age, creating a guaranteed RRIF is highly advisable.

In my view, GMWBs are longevity insurance. And while it seems a little ironic we would have to protect ourselves from long life, as mortality rates increase and retirement gets longer, it’s definitely something all advisors and clients should take into consideration. We don’t know how long any of us are going to be around, so it’s best to prepare for long life.