Wednesday, June 1, 2016

Pensionize Your Nest Egg

There are two central messages of the book Pensionize Your Nest Egg, by Moshe Milevsky and Alexandra Macqueen. The first is that having assets saved for retirement is not the same as having a pension. The second is that you can turn assets into a pension with the right insurance products. Both are true, but I have serious reservations about the variable annuities the authors recommend.

To a first approximation, the authors say that if there is a nontrivial chance that you’ll run out of money in retirement, then you don’t have a pension. Any attempt to manage your asset allocation through bucketing or other means can leave you vulnerable to the risks of living long, inflation, getting poor investment returns, or having poorer returns early in retirement that deplete your nest egg. The authors do a good job of explaining each of these risks.

The offered remedies for all of these problems are various types of annuities sold by insurance companies. The authors offer a process for determining how to split your nest egg across stocks, bonds, simple annuities, variable annuities, and various types of guaranteed income riders with these annuities. They call this “pensionizing” your nest egg.

There is no reason, in principle, why this can’t be great advice. Having an insurance company average out longevity risk across many retirees can benefit everyone. The problem I’ve seen is the sky high fees embedded in the complex annuity products I’ve examined. In one case, the guaranteed income looked good with the possibility of income increases if stocks performed well. But the rules for when these income increases would kick in combined with huge yearly fees made it very likely that few if any increases would occur. This would leave the retiree with income eroding with inflation each year.

The authors assert that “converting some of your initial nest egg into a stream of lifetime income by pensionizing it increases the amount you can spend at all ages, regardless of the exact cost of the pension annuity.” This is clearly false if the fees baked into annuities are too high. I can only assume the authors imagine the fees to be limited to reasonable levels.

The authors do compare the costs of annuities to the costs of other types of investments. However, the annual percentages given will mislead many readers. They say that the total annual fee of the typical Canadian mutual fund is 1.84%/year, and that variable annuities with lifetime income guarantees have annual fees in the 3-5% range.

This isn’t a fair comparison because some of the variable annuity fee exists to cover the cost of the income guarantee. Even ignoring this fact, to most people, the cost difference doesn’t look huge. However, suppose that the average dollar of your nest egg stays invested through 15 years of retirement. If we convert these annual costs to 15-year costs, the mutual fund costs 24%, and the variable annuity costs 37-54%! Even seemingly small increments in fees make a big difference.

The book’s table of costs also lists the fees for exchange-traded funds (ETFs). Strangely, though, it lists these fees as starting at 0.5%/year in Canada and the U.S. It’s not difficult at all to find index ETFs charging less than 0.1% per year.

Many commentators advocate a retirement strategy of owning more bonds and fewer stocks as you age and placing some fraction of your nest egg in a simple annuity when you reach 75-80 years old. As long as the stocks and bonds are invested in low-cost ETFs, I’m skeptical that the authors’ pensionizing strategy is superior. I’d have to see the numbers to be convinced otherwise.

19 comments:

  1. 3-5% ??!! There is no sound strategy that could employ an instrument that cost 3-5% annually ...

    ReplyDelete
    Replies
    1. @Anonymous: As I said, this 3-5% includes the premium for the income guarantee, but my reaction is still similar to yours. It's hard to see how you'll get anywhere with your nest egg leaking like a sieve.

      Delete
  2. I cannot see the usefulness of variable annuities anyway. They do not address any of the key retirement issues (longevity, inflation, sequence of returns, liquidity, legacy) any better than a self-made combo of regular diversified portfolio and plain traditional annuities. In addition, the best deal seems to lie with the simple traditional annuities - lifetime, non-indexed, non-guaranteed term. Their money's worth ratio is quite close to the fair value of 1.0.

    ReplyDelete
    Replies
    1. @CanadianInvestor: The guaranteed income riders on variable annuities offer longevity and sequence of return protections. However, the lack of inflation protection (in the VAs I've seen) limit the value of these other protections.

      I think it's possible to create a VA that offers value, but it would have to be much cheaper and structured somewhat differently from the VAs I've seen. The main potential of VAs over traditional annuities is the possibility of maintaining greater exposure to stocks while still getting longevity protection. But I'm not holding my breath waiting for insurance companies to make such improvements.

      Delete
  3. The best way to pensionize your retirement is by working for a firm that gives you a pension? (OK, that is being a Smart-Ass on my part)

    If you want a constant income there are other ways, as you point out, and I really have a skewed view on Insurance companies (i.e. I don't trust their numbers), but that is my own financial paranoia showing.

    Part of your retirement will be pensionized, with CPP :-) (yes being a smart-ass again).

    ReplyDelete
    Replies
    1. @Alan: It's true that CPP gives a good pension income base that takes some pressure off the worry of running out of money.

      Delete
  4. I am with the rest of you in that I can't see buying annuities given the high costs and good ETF alternatives available

    ReplyDelete
    Replies
    1. @Marko: I could see buying a simple life annuity with part of my money in my late 70s. I'd want at least 2% indexing if full inflation protection isn't available. However, variable annuities would have to improve a lot to attract me.

      Delete
    2. Mike Piper's latest blog entry discusses whether fixed indexed annuities are worth having. He says not.

      http://www.obliviousinvestor.com/annuities-with-fixed-cost-of-living-adjustments-dont-protect-against-inflation/

      Delete
    3. @Garth: I think his conclusion is different from this. He prefers annuities with genuine inflation adjustments over those with fixed indexing. So do I. But if I can't buy actual inflation protection, fixed indexing is better than no indexing for trying to smooth out consumption.

      Delete
  5. Warning: crude ballpark estimate, since accurate annuity math is hard.
    As a single investor, I need to plan for my nest egg to last to an optimistic(?) lifespan, say 35 years after an age-65 retirement. An annuity, diversifying across individuals, can plan to make my and everyone else's share last an average lifespan, 20 years instead of 35.
    If I invest for 35 years, my target rate of return needs to be X, whatever a sustainable inflation-adjusted principal preserving rate it (subject to some debate), plus approx 1/35=3%. If I invest for 20%, it's X+1/20=X+5%.
    So very roughly speaking, the annuity's longevity risk mitigating benefits are worth about 5-3%=2% annual return, so that's the pie I need to share with the insco's increased fees for it to make sense.
    Therefore 3-5% in total annual costs (for an annuity product) make sense iff I would otherwise invest in a high cost (1-3% per year) nonannuitized investment portfolio. Not likely for informed cost conscious DIY investors, but not out of line for unsophisticated retail investors.
    All sorts of stuff disregarded to get a quick ballpark estimate.

    ReplyDelete
    Replies
    1. @Martin: Interesting analysis. I agree that the VA route might make sense for someone who would otherwise invest in high-cost mutual funds, but not for the cost-conscious.

      Would you happen to be the same Martin Pergler who worked as a student at Bell-Northern Research doing some MAC programming in the late 80s or early 90s?

      Delete
    2. Yes, that's me. Since then became a mathematician, then a consultant and risk manager. Were you there too? Sorry your name is less unique :)

      Re VA's: I distrust things that are opaque and try to do too many things. So like you I see value in having a simple, fixed/indexed annuity as part of one's financial portfolio to mitigate longevity risk (and thank the authors for bringing that thinking out into the open). I have a harder time seeing the value of a VA over an appropriate mix of fixed/indexed annuity and nonannuitized market-invested portfolio -- unless you can get a powerful tax arbitrage benefit, or really value the simplicity of outsourcing (for a fee!) to one product vs managing a portfolio.

      Delete
  6. Update with slightly better math (a quick'n'dirty Monte Carlo simulation):

    If you withdraw 3.5% per year from a portfolio earning returns of 4.2% mean with a 9% stdev (normally distributed, and assume these are all inflation adjusted/real returns), you have a vanishingly small chance of running out in 20 years (average remaining retirement lifespan) and a decent chance of 7-8% of running out by age 100. If you layer on 3.5% per year of fees (a drag down on the 4.2%), you have a similar 7% chance of running out in 20 years. That chance would go to 70% by age 100, but that's where annuities have the advantage since if you reach 100, someone else will have expired on the golf course at 66 to even out.
    This is still simplified (and doesn't account for the variable part of a VA). The asymmetry of ruin means my non-Monte Carlo calculation underestimated the value of the annuity longevity protection, but still means a (V)A with 3-5% needs to be compared with high- rather than low-cost investment portfolios to make sense.

    The 4.2/9% come more or less from the PWL Capital Bortoletti-Kerzelo whitepaper on return expectations. The 3.5% is pulled out of my a** based on the perennial internet debate that a traditional 4% SWR is slightly too risky, and since it makes the numbers work nicely.

    ReplyDelete
    Replies
    1. @Martin: Thanks for the data. All this makes me yearn for a well-run tontine invested partially in stock indices. I could handle a somewhat variable income if longevity risk were taken away. I'd also be happy to take the risk of a sudden rise in average longevity because just about every other form of pension would blow up anyway.

      Delete
    2. It seems time for some disruption in the annuity industry. You guys ever think about starting your own insurance company focussing on annuities? I guess there is a lot of overhead dealing with legal and regulatory issues. But I suspect it would only take 100 or so people to band together and pool their risk. Some sort of not-for-profit coop or Vanguard like structure maybe.

      Delete
    3. @Greg: I'm way too lazy and risk averse for such an enterprise, but I'd be happy to cheer on anyone who tries. I think it would take far more than 100 people pooling risk, mainly because we wouldn't want anyone to derive significant benefit from bumping off other pool members. I think any well-run tontine should probably have rules about shutting down and distributing assets in the event that the number of participants drops below some threshold.

      Delete
  7. A monthly income stream based upon high-yield (7%) closed end funds is likely a better investment than an annuity; and you can get your principle back any time you desire.

    ReplyDelete
    Replies
    1. @Bob: The high-yield income funds I've seen look dangerous to me. They have unsustainable payouts. Many of them have had to cut their payouts after a period of declining capital value due to their income including return of capital.

      Delete