This is the fifth and final in a series of blogs on what advisors don’t know, but should, about variable annuities. Previous installments in the series have covered the mechanics of how Guaranteed Lifetime Withdrawal Benefits (GLWB) and Guaranteed Lifetime Income Benefits (GLIB) provisions work inside of variable annuities. In the most recent post, we discussed how current market conditions impact the benefits of VAs.
In the previous installments in this series we’ve taken a serious look at the mechanics behind variable annuities. We’ve tried to demonstrate that these are complex products with features and benefits that advisors need to weigh before they go down this path with a client. After taking all those factors into consideration, however, there are still valid reasons for advisors, under the right circumstances, with the proper metrics and allocation framework, to include a variable annuity in a client’s portfolio of investments.
One of the tricky parts of dealing with variable annuities for advisors is that not only do they have to ask the right questions, but the answers to those questions are likely to change depending on a large number of variables. All VAs are not the same, nor are the investor portfolios where they need to fit due to factors such as asset allocation, risk exposure, expense levels and how much income is needed.
Properly assessing the viability of a variable annuity for a particular client entails considerably more than just adding it alongside a standard model portfolio. Advisors instead should use a process that will allow them to calculate the optimal allocations toward guaranteed income products including modeling the difference between a specific guaranteed income product and a model, asset-allocated portfolio used for systematic withdrawal (SWiP).
There are three key factors for which advisors should examine when determining the potential success of any retirement plan, particularly those that include a guaranteed income product:
- Probability of Ruin
This is the probability that the retiree runs out of money and will not be able to fund their income goal while still alive. This probability depends on two random processes: the returns on the investor’s investment portfolio and the time of death.
- Expected Shortfall
This is the average, mortality-weighted, discounted amount across all Monte Carlo forecasts of the extra retirement wealth needed at the start of retirement to make a given investment strategy successful.
- Expected Death Benefit
This is the present value of the projected remainder of the retiree’s investable portfolio at the end of their mortality-weighted lifetime.
Run Scenarios in Monte Carlo
Given the complexity of the problem, using a Monte Carlo approach is preferable to an analytic method when looking at most guaranteed income products. It is also imperative that the advisor keep in mind that the income derived from both a guaranteed income product and a SWiP are dependent on the performance of the capital markets, and that the performance of one product/portfolio can influence the cash flows needed from the other. By using a Monte Carlo approach, the advisor can model the nuances of any guaranteed income product, look at how the capital markets impact their behavior and accurately capture the intertwined relationships of the possible combinations.
In the optimization process used by my company, Achaean, a Monte Carlo forecast is generated for a given set of client circumstances using each combination between the SWiP and the guaranteed income product. This starts by allocating 0% to the guaranteed income product and 100% to the SWiP portfolio and then ratchets up the increments in the allocation to the guaranteed income product in 5% increments (e.g. 95% SWiP/5% Annuity, 90% SWiP/10% Annuity, etc.).
The Probability of Ruin, Expected Shortfall and Expected Death Benefit are recorded for each allocation combination. These statistics are then reviewed to find the allocation that balances minimizing Expected Shortfall and Probability of Ruin while meeting an Expected Death Benefit constraint (e.g. 20% of initial wealth at death). The client circumstances that need to be included are age, gender, desired withdrawal rate and a portfolio risk level.