Our Pivot to a Liability Matching Portfolio
In this article I’m going to share my thinking on our pivot to a liability matching portfolio (LMP) combined with a risk portfolio (RP). A liability matching portfolio doesn’t get much attention in the press and blogosphere, so you may not know what it is. But, before I get into the “what”, let me cover a bit of our investing background and why we’re considering a change to our current strategy.
Like most fellow investors still in our working years, Jenny and I are accumulating wealth to fund our retirement. I have carried a portfolio of 80-100% equity mutual funds for nearly three decades. The equities have been spread among large and small cap, with about 25-30% exposure to international. I’ve generally leaned towards value and kept my growth exposure limited to what the S&P 500 provides. Our bond holdings have been short to intermediate government bond funds with some high quality corporates sprinkled in. I’ve mainly stuck to the typical total bond market index fund. Duration for those interested averages around 7 years.
A portfolio with that share of equities is considered an aggressive mix by many, particularly at our ages. The traditional advice of “your age in bonds” would have us at about 50% bonds. Until recently we were far short of that. But, I’ve always had a high risk tolerance and sense of accountability. Our portfolio has stayed the course through two steep declines (2000 and 2008). We’ve reaped the benefits of continuing to buy in the down years. My view has been with many earnings years still ahead, we can tolerate the worst (historically) the market can throw at us.
I would say that I enthusiastically signed up for all that the modern portfolio theory (MPT) and the efficient frontier had to offer. While I made plenty of missteps through the years, MPT and its focus on attempting to optimize a portfolio mix to maximize total return (suited to the individual investor’s risk tolerance) served us well. We have accumulated a nice nest egg and it continues to grow.
But, with age and a realization that some form of retirement is approaching faster than ever, comes a reevaluation of our investing strategy. Fortunately, my years of interest in investing and finance gave me much to consider when thinking about a strategy change.
At this point, we’re about 4-9 years away from a non-traditional retirement. I say “non-traditional”, because it is likely we’ll continue to work to some degree. When we exit the full-time labor force, our portfolio will need to provide the bulk of our income, but how will it do that?
There are many options and it would take a very lengthy article to cover the majority of them. Let me limit the discussion to the two major categories of retirement income that I considered. The first is the much publicized systematic withdrawal method (the “4% rule” and it’s many many variants, to include all manner of decision models, etc.). The second, is a safety-first option with a liability matching portfolio and a risk portfolio as one of the methods in that category.
Now, you may be thinking, “Didn’t he just say he has a high risk tolerance? Why is he considering a ‘safety-first’ strategy?”. Because, I’m very risk tolerant when I have a predictable income that is expected to last for 10+ years. In other words, when I still have plenty of human capital remaining, I am quite risk tolerant. If we fast forward 10 years from now, our human capital will unfortunately be waning and our ability to just wait out a market downturn while still earning money will be greatly diminished. That situation gives me lots of pause about risk. Risk means stress and uncertainty. And for us, that doesn’t sound like an aspect of retirement we wish to take on.
The Risk of the Systematic Withdrawal Strategy
So, what is it about systematic withdrawal methods that make them risky? The very engine that makes them work, unfortunately. The return you get on your investments is for taking on risk. Risk and return go hand-in-hand. Within reason, the riskiest investments have the most potential for high returns. That’s a good thing when time is on your side and you aren’t immediately depending on that accumulated portfolio to pay your electric bill.
Accumulation portfolios attempt (with some success) to smooth out the volatility and risk by diversifying amongst investments that have varying amounts of volatility (standard deviation) and non-correlation of their returns. This is why you typically see portfolios with mixes of stocks (equities), bonds, and cash, among other investments. The high flying stocks with their higher risks and returns are moderated by the lower risk (and lower performing) bonds. If one investment type is up while another is down (i.e. non-correlating), that smooths things out even more. Cash is there as a buffer for short-term expenses, primarily when markets are down. The cash also allows for “dry powder” to invest when markets are depressed and deals abound.
The risk of systematic withdrawal strategies on retirement lifestyle can be almost completely eliminated if the accumulated portfolio is large enough compared to the lifestyle income required. Many studies have shown that if you get the withdrawal percentage down below 3% of the portfolio value, you can weather protracted market doldrums for time periods well exceeding 30 years. This generally means that your retirement portfolio value at time of retirement would be at least 30 times more than the yearly income you need. So, for example, if after pensions, social security, etc. you still need $50k per year, you would need a portfolio valued at least $1.5M.
Understand that I’m trying to keep it simple. Of course, the analysis of any specific situation is far more complex than a simple multiplier, but it’s a good place to start to understand the magnitude of accumulated portfolio to consider. Unless your portfolio size is a multiple of your income far in excess of 30x your yearly retirement income need, the success depends greatly on you managing your withdrawals, closely monitoring your portfolio returns (at least yearly) and making adjustments as needed. Withdrawals greater than 3-4% may be possible, but not without significant consequences during down market years. This is particularly true if taken early in retirement when many retirees want to spend the most.
Sequence of Returns Risk
This problem of taking too large a withdrawal early in retirement and running out of money later is also aggravated by the “sequence of returns risk”. The order of the market returns during your retirement is critical. While we all want high returns throughout our retired years, a run of 30+ years has yet to happen in history. Instead, we can expect at least the same volatility we’ve seen for the last 8 decades. The problem is, if the down years occur in the first decade of a retirement it can have a devastating impact on a portfolio’s ability to support desired income for the long term. Most every systematic withdrawal strategy beyond a simple percentage withdrawal looks to minimize this risk by lowering income when the market is down. Maintaining your lifestyle becomes critically dependent on managing your portfolio and the level of systematic withdrawals. This requires analysis and intervention – each and every year.
What has me shying away from that approach is the risk to maintaining a predictable and desired income in our retired years. Our multiple is right around that 30x discussed above, which means a retirement 40 years long could be a nail biter if an unfavorable sequence of returns rears its ugly head. Additionally, the maintenance required to ensure we’re not overspending or that our portfolio is well allocated for the years ahead has me concerned. Why? Cognitive function, that’s why.
We take cognitive function for granted for much of our lives. The reality is that some forms of cognitive function, particularly math and numeric skills, are believed to decline as we age. This rate of decline would definitely vary among the population. And, this is not to say we are no longer cognitively competent beyond a certain age, but we must be realistic about depending on a robust analytical mind in our later years to make decisions about our finances. I am uncomfortable devising a retirement income strategy that requires me to be very technically competent in my old age for it to be successful. On top of this, if I die before Jenny (a likely event since she is 4 years my junior), she’ll be left with a non-simple plan to execute. Either way, I would likely need to hire a financial consultant to help us manage the plan into old age. I’d like to avoid that as long as possible.
If I’m not feeling comfortable with the options I’ve explored with accumulation portfolios and systematic withdrawals, then what do we do? My research led me to the liability matching portfolio with a risk portfolio.
Liability Matching Portfolio with Risk Portfolio
What is a liability matching portfolio? It is an investment portfolio, or a portion thereof, the assets of which are used to offset a planned liability in retirement. A LMP is most often used in conjunction with a safety-first strategy mentioned above. The LMP is “matched” to the desired income needs in retirement (the liability). “Desired” can mean many things. To some it may be the bare minimum to survive – basic housing, food, clothing, and medical care. To others it could be income to fund an extravagant lifestyle. And…everything in between. The more desire you have, the bigger your LMP will need to be. That is no different than the accumulated portfolio with systematic withdrawals. The more lifestyle you want it to support, the bigger it needs to be. There’s simply no free lunch.
The difference between a safety-first implemented LMP and an accumulated portfolio with systematic withdrawals is that word “safety”. The LMP in this case is intended to be converted to a sure thing level of income via annuities, longevity insurance, bond ladders, or other methods that generate guaranteed income. This “income floor” can be implemented with or without inflation protection. The tradeoff is that for lower risk, you get less return. This means you would need more assets to cover that floor than you would likely need for systematic withdrawals. BUT, if structured properly, there is very little risk to income.
The remainder of the accumulated portfolio (the Risk Portfolio) is invested as aggressively as the retiree wishes, to provide an up-side to retirement income. Since the floor income needs are already provided by the liability matching portfolio, the retiree can potentially feel less stress about the risks involved with the RP. The RP can be allowed to grow over time to provide for the possibility of long-term care, or as a bequest to heirs or to a social cause.
Comparing Systematic Withdrawals to Liability Matching Portfolio with Risk Portfolio
I’ve summarized my comparison of the two approaches in the table below.
|Accumulation Portfolio with Systematic Withdrawals||Liability Matching Portfolio with Risk Portfolio|
|Portfolio Maintenance||More Complex||Less Complex|
|Inflation Protection||Depends on Investments||Depends on Investments|
Systematic withdrawal strategies will generally have the potential to produce a higher level of income than a liability matching portfolio strategy with floor, particularly if inflation protection is included. How much more income is dependent on the level of risk taken by the systematic withdrawal portfolio and on the other side how much risk is taken by the risk portfolio portion of the LMP/RP strategy. My analysis has shown that for our situation, the size of the RP is not large enough to overcome the return advantage of the larger accumulation portfolio used with systematic withdrawals.
While not trivial to set up (purchasing annuities or building a bond ladder), the liability matching portfolio floor is much simpler to maintain than the portfolio used for systematic withdrawals. Depending on how complex a person wanted their risk portfolio to be, it could be just as complex to manage as the systematic withdrawal portfolio, but since the stakes are lower for the RP, there is less stress about optimizing it. And, depending on the purpose of the RP, a set it and forget it allocation might be all that is needed.
I think systematic withdrawals have an edge on income flexibility. Since many of the LMP strategies lock-in an income via an annuity or bond ladder, there can be less flexibility in varying income from year to year. The situation is improved if the LMP floor is structured with bond ladders only or with a mix of annuities and bond ladders. Future rungs of the bond ladder could be spent early to increase current income and replaced with proceeds from the risk portfolio if available. Alternatively, the RP could be mined for additional income if required. Managing these two portfolios for unexpected needs seems more complicated to me than one overall portfolio used for systematic withdrawals. The solution is to have a larger floor to cover unexpected expenses, or use the RP as a shock absorber and mine it only lightly for discretionary expenses.
I judge the two strategies as equivalent in their ability to guard lifestyle against inflation. Systematic withdrawal portfolios usually have a non-trivial stake in equities and a portion of their bond holdings could be satisfied by treasury inflations protected securities (TIPS). Likewise, the liability matched portfolio floor can be structured using a TIPS ladder or annuities with an inflation rider. While the risk portfolio of a LMP/RP strategy can provide inflation protection via its equity holdings, the RP is probably not big enough to protect the combined LMP/RP.
This has been a long article, but our shift from solely an accumulation portfolio to a liability matching portfolio was a big decision. The decision came down to two personal preferences for us:
- The need for a guaranteed income floor, even if the level of income is lower than could be attained with a riskier strategy.
- The need for an easy to implement strategy to make aging less stressful and easier on Jenny should I not be around to manage our finances.
I fully recognize that this strategy may not be for everyone, particularly if you are determined to grow your retirement portfolio as large as possible to support income or bequests and you are willing to accept the risk and complexity that goes with it.
In a future article, I will present a generalized sample LMP/RP approach as I intend to implement it for our retirement. Stay tuned.
Posted by Jeff