Panelists at Markets Group’s Sixth Annual Texas Institutional Forum, in Austin, gathered together to tackle the pros and cons of implementing obligation-based or liability-driven investment (LDI) policies at pension funds. The discussion was predicated on the idea that the typically complicated process of administering the investments of a pension plan can be simplified, somewhat, by focusing on a fund’s obligations or liabilities. The approach is built on two very hard to dispute assumptions: Most pension plans have very well-defined and predictable, required pension payments, and the long-term nature of pension obligations allows pension investments to have very long investment horizons. Though the panelists differed, somewhat, on the finer points and nuances of the issues being discussed, on one question there was agreement: the continuing closure of defined benefit plans, due to the depletion of funding, is both lamentable and avoidable, and large factor in a potential retirement income crisis.
The panelists included Moderator Eileen Neill, a managing director and senior consultant at Verus, David Wilson, a managing director and head of taxable fixed income, client portfolio management at Nuveen Asset Management, Christopher Shelling, director, private equity, Texas Municipal Retirement System, Colin Kerwin, retired global pension fund manager, ExxonMobil, and Jeffery Blazek, managing director, Cambridge Associates.
Neill observed that when she joined this industry in the early 1990s the tech bubble had just started building. At that time, portfolios were returning double digits—equities and bonds. Because assets were performing so well, little attention was given to the liabilities, she said, recalling attending client meetings and quarterly meetings that were generally short and the lunches long.
“And that occurred until 2000 when the tech bubble burst, followed by Enron, followed by Long Term Capital Management, followed by the subprime crises, and then, ultimately followed by the global financial crisis,” Neill said. “During that period—the early 2000s—we all woke up and said ‘Oh, we’ve been focused on the assets side and here are liabilities that have grown, and we’re all underfunded. And the regulators came in, and they implemented the PPA [Pension Protection Act] in 2006. So, that really was a game changer for corporate defined benefit plans. They started to think about structuring their plans very differently than the public plans.”
Neil began the discussion by asking panelists how they each think about structuring their portfolios, in terms of incorporating the effect of liabilities for which they have oversight.”
David Wilson: I work with corporate pensions, endowments and foundations and public pension funds and insurers. All have unique liabilities. In the case of defined benefit plans, employees accrue benefits that are going to be paid out at retirement. Now, accounting rules have made corporate and public pensions very different. So, for example, with a corporate pension plan, you are still paying out cash flows for the life of the retiree. But, that liability has to be marked to market using high-quality corporate bond yields. So, if you want to defease that liability, or mitigate the funded-ratio volatility, corporate bonds that match the duration of the liability is your best tool. If you are talking about public-plan liabilities, there is no real duration because the liabilities are discounted using a relatively static discount rate. So, if you want to hedge that liability, cash-flow matching is the preferred approach.
By contrast, endowments and foundations have spending rules. Let’s say their spending rule is five percent. Well, that’s an obligation that has to be paid every year; they need to construct portfolios that minimize the volatility—or I should say, maximize the probability of meeting that spending rule without depleting the portfolio. So, very different approaches depending on the accounting and regulatory environment.
Colin Kerwin: For me, I think it starts with understanding first and foremost what is the purpose of a pension plan? Is it to pay for benefits that are being accrued, or is it to secure the benefits that have already been accrued? I would argue that it is the latter. Certainly, under ERISA, the purpose of the plan is to safeguard qualified retirement benefits. The funded plan assets are there to secure those obligations. Separately, it’s the responsibility of the sponsoring organization to fund those obligations. So, if the purpose is to secure the obligation, what is the purpose of risk taking inside the pension plan or, said differently, how does a risk mismatch between the assets and the liabilities help to safeguard existing benefits? I don’t think it can. I would argue that if the purpose of the plan is to safeguard the obligation then the goal should be better matching of the risk characteristics between the obligation and the assets.
So, for pension plans, that should mean very long duration, high-quality, fixed income on the investment side. From a plan participant’s perspective, they arguably own what should be a risk-free promise—one backed by both the sponsoring organization and the funded plan assets. So, a portfolio of long-duration, high quality fixed-income that closely matches the duration and convexity of those liabilities will effectively track the liability, no matter what happens to interest rates, no matter what happens in the financial markets—thereby safeguarding it.
The last observation I would make is that this approach should be a win-win for both the plan participants and the sponsoring organization. From the plan participants’ perspective, they are better off if the assets are managed this way, because the assets are collateral for a promise made to them. From the sponsoring organization’s perspective, they too should be better off because, in most cases, the business of the sponsoring organization is not to take risks in financial assets— shareholders (and taxpayers on the public side) can already do that for themselves.
Christopher Schelling: I still am employed with the Texas Municipal Retirement System (audience laughter) so I will try to split the difference between the house view and some of my own thoughts. Colin makes an excellent point on securing retirement versus providing retirement. Our actual mission statement is to provide a secured retirement—so we have to do a little bit of both.
I’ll take a step back and maybe differentiate between Texas Municipal and a typical old, state defined benefit plan. Generally, in a state defined benefit plan, you have a legislature that will set the contribution rate and will determine what the benefits are. The benefits are often constitutionally guaranteed. You then have a target rate of return and an asset management division that manages those.
Texas Municipal, seventy years ago, was set up to be a cash-balance hybrid plan. So, it was very innovative in that regard. But, we don’t receive funding from the state. We are a quasi-agency; we don’t have a budget line item in the state’s budget. All of our funding comes from individual cities that are voluntary participants in our plan. So, we have different benefit levels that our board gets to determine and offer those packages to our constituent cities. And we require different funding levels for those different benefit packages. And our board actually gets to set on an annual basis what the contribution rates are for our member plans, and gets to determine what the required rate of return is. So, we think of it as a slide rule—we kind of have control over both sides of the slide rule, which helps us smooth out those changes over time, and weigh the tradeoff between generational wealth creation and wealth transfer, so to speak. So, we kind of view—as the investment team—our job as to hit the rate of return that the board sets up. And we certainly can opine on some of the cash-flow characteristics of the short-term obligations, and think about those in the construction of our portfolio.
Jeffrey Blazek: I think one of the interesting things about my role is that I get to work with a variety of clients. And the way that we think about the liability is that it is driven by a very simple question: Is the plan still open or not? And the trend has been that many are closing. And I think that is tragic for a lot of reasons that we might touch on in this panel. But that’s not our choice—we just take the orders. Where does the plan stand? And if it is closed, then we look a lot at the liability. We want to reduce balance-sheet risk and the funded-status risk as much as humanly possible. PPA changed the rules on that—it may have well been called the bond manager employment act! I mean, it was going to drive a huge boon of assets to long-duration strategies. Because that is the largest risk that a hard-frozen or soft-frozen plan will face. And so, we are going to invest that portfolio thoughtfully, but it’s going to be very focused on the liability.
But my former employer, New York Presbyterian Hospital, still has an open and accruing plan. So, when I was in-house at that investment office, we invested it identically to the endowment because the way we looked at it from the hospital’s perspective was a dollar was a dollar, and if we had a shortfall we were going to own it one way or another as a plan sponsor. So, we may as well seek to maximize income with this open and accruing plan. So, at Cambridge, when we have open plans, we do the same. We take some liquidity risk, we maximize equity, thoughtfully. We look at the liability. We want to make sure that we put the sponsor in a position to succeed and really make up return to service costs that they are going to face in the future.
Partially as a result of PPA, there have been a lot of corporate DB plans that have been shut down. And even in public fund land there have been funds that have been frozen and replaced by defined contribution plans. I think the issue is are we opening up Pandora’s Box for a different kind of crisis than the perceived crisis of the defined benefit plan by forcing participants to invest in defined contribution plans.
Neill: That’s a good segue into the fact that investors’ time horizons do have an influence, in terms of their ability to take risk. And with many public pension plans today, with bond yields so low and the equity risk premium lower than historically is the case, we are seeing kind of a pedal-to-the-metal investment approach occurring across the board. The sentiment is we have this very, very long time horizon. But, I know that there is some difference of opinion on the panel about whether or not a very long time horizon should drive or influence how you structure the investment program. So, let’s talk a little bit about what are some of the pitfalls of that approach.
Kerwin: For me, I like to think about the return inherent in the liability as the starting point. So, put yourself in the shoes of a public plan participant. You are promised a lifelong defined benefit from your employer when you retire. You accrue these benefits during your working career by accepting lower current compensation (there is no free lunch). So, ask yourself, what is the return I’m getting on this deferred compensation? Another way to think about it is the plan participant is making a loan to their employer (by accepting a lower cash salary) and will be repaid with interest when they retire in the form of a lifelong annuity. What is the appropriate interest rate on that loan? Well, given the loan in theory is fully collateralized with pension assets and further guaranteed by a sponsoring organization often with taxing authority, the return should be close to a risk-free rate. So, regardless of accounting or actuarial standards, the economics are that the returns inherent in a public defined benefit cash flow stream should be close if not equal to a risk-free return.
If the return is risk free, then I would argue that the only return that trustees need to generate to keep pace with the economic growth in that liability through time is a risk-free return, which can be easily generated in a simple low-cost fixed-income portfolio. As soon as the trustees introduce a large risk mismatch, in either an attempt to close a funding gap or to meet future pension obligations, then not only are they undermining the safety of the existing promises, but they are also transferring wealth from future taxpayers to current taxpayers. It’s like saying ‘I’m not going to fully fund the bill for services already delivered. Instead, I’m going to rely on the financial markets and long-term risk premia to fund the balance.’ But it’s the future taxpayers that are bearing the risk of that choice. And if it all works out, and the fund grows, and obligations are satisfied, then the future tax payers that took all the risk never receive the benefit because the current taxpayers have already claimed it by underfunding the plan. So, I think introducing a risk mismatch, particularly in public plans, is really nothing more than a wealth transfer from future generations to the current, under the guise of intelligent investing.
Neill: Chris, when we had a prior discussion I thought you had some really good points particularly given perspective as a very long-term investor in private equity.
Schelling: Sure, I have a slightly different view. I would point to a kind of an analogy: In a defined contribution plan there’s a twenty-something-year-old investor. You would be considered grossly in violation of your fiduciary duties if you told an investor to put 100% of their 401k in bonds. So, there’s got to be some kind of trade off from value creation from the investment returns versus taxing the public, and I’m not claiming that generational wealth transfers in the public pension have been effectively managed. Certainly, we have room to improve as a sector. If you compare some our governance and processes versus the Canadians or Brits, Nordic or Australian plans, they’ve done a different job. I think that part of that comes from explicitly managing both sides of the balance sheet—liabilities plus assets together—and when you look at some plans that have done it right—I could point to the State of Wisconsin—they have different assumed rates of return and discount rates. So, they are discounting at a lower rate and targeting a slightly higher rate of return. And that’s kind of a margin of safety as opposed to introducing too much risk. You should be saving more, but if you can make the assets work harder for you in the long run, then I think it is your responsibility to do that. And that is reflected sort of in our mission statement, which it is to provide that.
We are a little bit different as a cash-balance hybrid plan. So our benefits payments tend to be a little bit lower than a true DB, and our contribution rates are a little higher than a true DB. And we do have some market forces at play, because we have voluntary participation from all of our constituent cities. So, we have to offer them a package that is attractive for them to participate in. The flip side of that is we can’t charge too much for it, otherwise they wouldn’t participate.
But we can also grow. We are healthy from an active to retiree participant perspective, because we are able to add new cities every year. It is truly a hybrid plan, but that doesn’t flow into our asset allocation. When you look at a cash-flow perspective, we have a very low spending rate. It just turned negative for the first time ever, and it won’t reach above 2% for a couple of decades out. So, I think you really can invest that portfolio markedly different from one that has a vastly higher percentage of retirees to active participants and different cash-flow characteristics.
So, there is a generational inequality in the population that is wreaking havoc on any promised obligation, whether it is Social Security, Medicare or defined benefit pension plans.
Wilson: I think there are two issues at play here. Let’s talk about demographics first—the demographics of our country. Baby Boomers, people born between 1946 to 1964, 80 million strong; Generation X, people born between1964 to 1982, roughly 65 million—so 15 million people less. And then you add millennials, that’s approximately 85 million people. So, there is a generational inequality in the population that is wreaking havoc on any promised obligation, whether it is Social Security, Medicare or defined benefit pension plans. So, when you talk about time horizon, think about the demographics of your plan, there are some hard numbers that can give you guidance on what your investment time horizon really is. If you look at the active-to-retiree ratio in your plan, and you look at the trend, it’s getting close to one-to-one—one active employee to one retiree. If you look at the net cash outflow of your plan as a percentage of the asset base, that’s going more and more negative—your investment time horizon is shrinking.
We always think about pension funds as having this really long investment time horizon. But that isn’t the case anymore. As the baby boomers retire, pension cash flows go up—the obligation goes up for the next 20 to 25 years, and that’s causing a lot of deficits with pension plans in our country. Now, how do you discount that liability?
So, I disagree that that’s a risk-free liability. I think the pensioners of the City of Detroit and other places would say ‘that wasn’t too risk free.’ Yes, many are guaranteed by state constitutions. But until they are tested—it was assumed that the City of Detroit’s pension plan was guaranteed until their Supreme Court ruled it as a contract not a guarantee. So, we would have to test these to really truly say they are risk free. I would say they are collateralized or secured, therefore the discount rate for the liabilities should be lower than seven and a half percent. Then there is the reality of the situation. If you discount public pension plan liabilities at the risk-free rate, the aggregate unfunded liability would increase from about $1.5 trillion to about $5 trillion now—you would have a massive debt crisis on your hands. So, I don’t think the liability is risk free, but I do think that the liability is discounted at a higher rate than it should be.
Blazek: I think that the discounting mechanism is important. If you do have a new person who starts working as a teacher, and they’re in their twenties, and you know they are not going to retire for another 30 years, you have to use that time to your advantage if you are the [plan] sponsor. If there are 30 years before they will start drawing cash from you, and you are a believer in the equity risk premium, then you would not be serving either the participant or the state well if you did not invest in equities to some extent. You have to be thoughtful about how you take that risk and monitor it—there’s a lot of complexity to that. I think the overall problem though is that these plans, particularly at the state level, have not been well-funded at all. It’s been about the willingness of political forces to put cash in, and the plans I work with, which are largely not in the public realm anymore—hospitals, non-profits, corporations—have been told very clearly for about 10 years now these are the rules, this is how you will be funded. It’s not been a lot of fun, but we have managed through it, and we have an average funding level of 90 percent on ERISA plans, despite a very choppy trading environment. And yes, that has caused a lot of them to close down, and that’s the tragedy of it—that’s how they’ve tried to solve the problem. But we are not going to have people that are going to have to eat [suffer] less benefits than they were promised. We’ve solved the math. I think the public realm has a real problem where they have simply not put enough money into these plans.
Neill: I think part of the difficulty is how we are actually valuing the liabilities of the plans, and what truly is the level of underfunding or fully funding? On the corporate DB side, the regulations basically force corporations to discount liabilities at a corporate bond rate—it treats liabilities as a bond. Whereas on the public fund side, it’s a little bit circular, but the discount rate is a function of your asset allocation. So, I’d be interested in hearing panelists’ views on that, given that you have a regulatory framework, and given that you have an investment framework that drives the value of your liabilities, how can you use that to potentially structure the portfolio, so you maximize the benefit or growth of those assets, while still working within the framework of that regulatory environment or practice paradigm?
Blazek: It’s going to depend on how large the pension is relative to the balance sheet and the overall risk level of the enterprise. When we do see things like the PPA imposed and the way that liabilities are measured, whether we agree with it or not, that’s the reality and that affects their funding level. The accountants have taken away smoothing for their financial statements. So, we used to have the ability to kind of amortize shortfall, but they basically mark-to-market now. And it’s a risk they think about daily. And they agonize over a 30-year move of 20 basis points. So it does influence the way we invest the portfolio, because if we don’t manage that risk effectively their debt could get downgraded or they may not have cash for their core operation.
I think introducing a risk mismatch, particularly in public plans, is really nothing more than a wealth transfer from future generations to the current under the guise of intelligent investing.
Kerwin: I guess it’s not going to surprise you that I see it a little differently. For the last 34 years I’ve practiced corporate finance at ExxonMobil. And one of the things I was taught at business school was that if you have a cash-flow profile in an open marketplace, that cash-flow profile can have only one value from an economic perspective. Otherwise it would simply be arbitraged away. Let’s assume you have two identical defined benefit plans, one in the private sector and one in the public sector. Because the cash-flow profile is exactly the same in both cases, there is no way on earth that those two streams should have different values. I don’t care what the accounting or actuarial standards may say, from an economic perspective, they have to have the same value. And I guess if they have the same economic value and they have the same risk profile, the investments you manage against them should also be the same. And they should also be the same whether the plan is open or closed. As I said earlier, it is not the responsibility of the pension plan to pay for that future liability. It’s the responsibility of the sponsor or, in the case of a public fund, the taxpayer, to pay for that additional liability. You can’t ask the asset to work overtime without creating a risk mismatch and getting into trouble.
Keep in mind that for every $1 of pension benefits, there are only two ways you can have an underfunded plan: 1. You didn’t fund it fully in the first place, which is a big part of the problem we see. Or you did fund it and then you lost the money relative to the liability because you had a risk mismatch between the asset and the liabilities. Both things are happening. Both things don’t need to happen.
Wilson: I agree with many elements of that approach. But the one thing it is ignoring is time horizon. If you breakdown a pension liability, it has three components: retirees, vested terminated employees and active employees. So the essence of demographic-based investing at Nuveen is to align investments with the characteristics of the liabilities. The retiree liability has more certain cash flows and is shorter in tenor. One can match this liability with assets that provide high cash flows like fixed income or certain alternatives. For active employees, where there’s an inflation component due to wage growth, that’s an uncertain cash flow with an uninvestable characteristic, namely how long people live. That liability needs an asset that provides growth. And it aligns with a longer time horizon using growth assets like equities. So, I think you can incorporate growth assets into your investment program, but you have to really consider the liability components of your plan.
Schelling: I think Colin raises an interesting point on corporate finance versus investment finance. What I would point to is in the corporate sector the pensions are inarguably healthier. They are 90% funded on average, with a 4 percent discount rate. In the public sector, call it 60-70 percent funded on average with a 7.5 percent discount rate. So, they are inarguably in a better position, better hedged against those long-term liabilities. I would say there may be an arbitrage that does occur with that long-term capital. And you see it actually in LDI [Liability-driven investing]. So, they fund up, and then they sell a big portion of those liability streams to insurance companies. And those insurance companies can earn more by investing in private equity. So, I’m saying there is a premium that can be harvested by being long term and investing in private markets, that’s kind of the take that we’ve had. We know our cash flow needs with a relatively high degree of certainty, and we can project what our cash flow from income and amortization of principal, etcetera is, and we are trying to earn an excess return.
Now, I would agree that there is governance at institutions that doesn’t necessarily allow that to be implemented effectively and that’s where we should address some of the problems in the public sector. There is a scale that is required. There are resources that are required in order to effectively access that premium. But that’s how we’ve taken it. We’ve moved the portfolio from the traditional 60/40, and even prior to that 100 % fixed income, to more 50/50 equity/fixed income. So we are probably more fixed-income oriented than many of our peers, but we’ve got a lot of that exposure through private markets.
Neill: Partially as a result of PPA, there have been a lot of corporate DB plans that have been shut down. And even in public fund land there have been funds that have been frozen and replaced by defined contribution plans. I think the issue is are we opening up Pandora’s Box for a different kind of crisis than the perceived crisis of the defined benefit plan, by forcing participants to invest in defined contribution plans.
Blazek: I think there is tremendous value in operating a DB plan. And if you treat it as a high-quality fixed-income instrument and quantify it on day one, that’s worth so much. To our teachers in Texas, and all throughout the country, that benefit is worth tremendous value. And if we see them converted to defined contribution plans, then they are getting short changed. Either their salaries have to go up to make up for the loss of that substantial benefit or the DC plans need to be structured in such a way that they are not taking on more of the burden. I think that conversation is such a dramatic conversion—we’ve seen it in the private market to the greatest extent. The salaries for public employees are often much lower, and it’s because they do get a very generous retirement benefit. That should not go away without some commensurate return that they get.
Wilson: I think pensions are freezing and closing their defined benefit plan at the very worst time, due to the demographic issue. It dramatically accelerates the shortening of your investment time horizon at the same time you are underfunded. So it puts a really big burden on the sponsor. If it’s in the public space, it puts a burden on the taxpayer. And it creates a future retirement problem for individuals with no acceptable solution in the defined contribution space, in my mind. Target date funds are good for a very novice investor who doesn’t have the time or capability to understand investment. But if you look at target date funds, and you look at the risk profile that they have when a person is nearing retirement, it’s unacceptable. There is too much capital at risk before retirement. We saw this during the financial crisis. For example, most 2010 target date funds lost 20 – 30 percent of their value in 2008. Now think about that if you are about retire and you are in a 2010 target date fund. How scary would that be? And guess what happened? A lot of people sold out of them and crystallized that loss. So, if you’re going to terminate defined benefit plans, which I would argue has an extraordinarily large social good to them, you also have to present a viable substitute retirement plan, which we have not done.
Kerwin: I’ve recently retired from ExxonMobil after 34 years of service. I could be out playing golf with my son right now (audience chuckles), but I feel really strongly about the defined benefit system. I was fortunate to work for a company that not only offered a defined benefit plan but continues to offer them for new hires around the world. I want to see the next generation of workers and the generations after that enjoy the same kind of retirement income security that I am now enjoying. And there is no reason at all that it can’t be done. If the defined benefit system is managed properly, it is no more costly than any other retirement system. But it is one that can provide secure lifetime income to potentially millions of people that don’t have the financial skills to manage their own portfolios. So, given the opportunity to speak on panels like this one and to defend the defined benefits system and to say there is a simpler, cheaper way to manage them, I’m going to come every time.
Schelling: On that we can completely agree. I do think the hybrid plan, as an intermediate step, is a little bit more salable. It may give a little bit less, but it accomplishes a lot of the benefits of a pension system in that it allows you to pool assets, invest for the long-term premium, etcetera. But it takes the decumulation problem off the beneficiaries, by actually monetizing it for them. We take our obligation to provide a secure retirement seriously and to make that available in perpetuity.