Pension Derisking Comes of Age

FRANÇOIS PELLERIN: There's been an acceleration of derisking this year because of tax reform. Pension plan contributions are tax deductible and if you make contributions before Sept. 15, you essentially benefit from the higher tax rates that were in effect before the tax reform bill became law. When you make contributions, your funded ratio improves and the logic of derisking applies.

JAMES SO: That's right. And part of that is also to reduce what they owe in [Pension Benefit Guaranty Corp.] premiums. The PBGC premium — the variable-rate component — continues to increase to the extent that a pension plan is underfunded. So when you make a contribution, you're getting a couple of different benefits. One is on the tax front, as François mentioned, the other is to reduce PGBC premiums, and then there is maybe a small income-statement boost.

GARY VEERMAN: There certainly have been some catalysts such as tax benefits and PBGC premiums, but the core reasons for derisking really haven't changed. It's about companies limiting the impact of the pension plan on company financials, creating certainty around potential cash contributions and satisfying benefit payments, but also, importantly, getting back to their core businesses.

GARY VEERMAN: It's really hard for plan sponsors to ignore the market environment. Some will argue that spreads are tight and rates are low, but it's easy to counter that with the point that we're at an all-time high in equities after an eight-year bull market. When you think about the components, spreads and rates of long corporate bonds are typically risk-reducing, while the equity portfolio is used to create excess returns beyond liability growth. So we would say you can add quite a bit of downside risk by delaying derisking at this point.

We believe that when deciding to implement an LDI program, plan sponsors should separate strategic asset allocation decisions and the current market environment. They should remember that LDI is not an all-or-nothing decision. It's a strategic asset allocation decision in which liabilities are a key input. If you believe this, some form of LDI should be implemented regardless of the market environment.

FRANÇOIS PELLERIN: Many sponsors think that rates are going to go up, thereby erasing a portion of their pension deficit. But what the [Federal Reserve] does on the front end doesn't affect the value of the liabilities by itself. Liabilities are impacted by the 10, 20, 30 year rates, and we don't think those rates are going to go up by much.

One important point about current low spread levels is that if spreads go up significantly, liabilities could go down by a lot as a result of higher discount rates. But if those spreads go up because of credit events, chances are that a plan's equity portfolio also gets decimated at the same time.

GARY VEERMAN: Strategic asset allocation is really what drives overall outcomes. Certainly, you want to put the decision making around markets in your manager's hands, whether that's equity managers or fixed-income managers.

Most plan sponsors are using stress testing and more traditional asset-liability analysis because, ultimately, there's a pain threshold for every sponsor that at some point additional downside in terms of funded status is too much for a company to handle. It's great to have that starting point, because then it's really about both the return-generating portfolio and the liability-hedging portfolio and how they fit together.

JAMES SO: LDI is a way of investing, and it means that the risks that you're taking are all relative to your liability. It doesn't mean that you can't be a long-bond investor if you're only 75% funded.

Those factors — the level of fundedness, the ability to take risk, the desire to do a pension-risk transfer where you sell the liability off to an insurance company — those are all part of the LDI discussion, but LDI just means being aware that you're investing to meet a certain liability stream. It's about being aware of how your assets move vs. your liabilities.

A colleague of mine used to manage the pension of a large international energy company and he said that 25 years ago, if a quarterly pension report for the investment committee was 25 pages long, the first 24 pages were about what the capital markets did and how the asset managers performed, and the last page was a report on the liability of the pension plan. Today it's the opposite.

FRANÇOIS PELLERIN: For a plan that is 85% funded, a simple actuarial analysis can demonstrate that markets by themselves are unlikely to get that plan where it needs to be, especially when including PBGC premiums. Such analyses help show sponsors the need to integrate investment and contribution policies. Once sponsors understand that, they can start planning for contributions. By derisking as contributions come in, the 'cone of outcome uncertainty' can be ring-fenced — how much it's going to cost to get to the end goal. So it's good old financial planning: Look at your risk tolerance, look at your budget, look at your investment horizon and design an asset allocation that will optimize that.

JAMES SO: Ten years ago, the first stage was to extend duration, or move from a core-type of fixed-income portfolio to long duration. Today they're diving in a little bit more. Glidepath design, completion management and asking for help in monitoring the funded status and making asset allocation adjustments as they move along the glidepath.

Another big thing recently has been customized benchmarks on the fixed income side. Over the last six years, we've seen a lot more of our clients want to slightly customize their benchmarks.

More clients today have some type of stated goal. In the past it was get as much risk-adjusted return as you could, because they were basically total-return investors. But today, a lot of pensions have some type of goal. Either it's getting to a certain funded level so they can hibernate the plan, move to a certain asset allocation that they think is going to run smoothly against the liability or get to a certain funded level where they can sell the liability to an insurance company.

GARY VEERMAN: The initial core-to-long move was a simple step that didn't impact long-term return expectations. It reduced risk. Most core bond managers could manage long-duration bonds, so it was an easy transition.

LDI used to be an exercise of buying long bonds. Today, I think LDI is really a multi-dimensional asset allocation exercise. The question we were asking before was, can you take my five-year bond portfolio and make it 13 years? It was as simple as that. The questions that we're asking today are things like, how should the LDI portfolio look given the size and composition of the return-generating portfolio? These are high-level asset allocation discussions. Questions like, what types of managers are appropriate as I build and expand my LDI portfolio, would not have been discussed when a plan sponsor moved a core-bond manager to a long duration manager.

Now the competitive landscape is changing. There are different ways to approach investing in bonds. How much interest rate should I hedge? No one really was asking that from core-to-long, they were just accepting whatever the outcome was. What should the portfolio look like at 100% funded? That's a great question to ask if you're in that position.

It goes on and on, which I think is what makes conversations so challenging but so exciting.

FRANÇOIS PELLERIN: Early on, sponsors went for the LDI path of least resistance: turning short-bond mandates into long-bond mandates. For a defined benefit pension investor, short bonds are a return-free, risky asset class. Long bonds are more efficient as they offer a higher yield and better funded ratio protection. Once that low-hanging fruit is addressed, other derisking strategies can be explored.

A common question is, when and how should I customize my LDI program? We say that if your plan is 80% funded and rightfully invested in a 60/40 portfolio, blunt hedging instruments may do the trick most of the time. But many of our clients' plans are close to 100% funded, so they need to derisk with more minutia. That's when tight key duration matching may add significant value. While derivatives can help enhance hedging, high levels of customization can generally be achieved with physical bonds only.

FRANÇOIS PELLERIN: We like to have derivatives authority in the investment guidelines to maximize flexibility. But the majority of our clients do not need derivatives to implement the right solution.

The classic example where derivatives would be justified is with a client that wants to carve out as much of the interest rate risk as possible because they believe — as we do — that interest rate risk is not compensated, and they want to make room for more return-seeking risk, which is compensated.

But plan sponsors have to be careful with derivatives mainly for one reason: The liabilities that people care about are typically the liabilities used for accounting, for determining contributions or determining the price to pay to do a pension risk transfer. All these liabilities are based on credit curves, so if you hedge with swaps or futures, you don't have a spread component and that can create surprising tracking error.

GARY VEERMAN: Plan sponsors need to know that derivatives are not a silver-bullet solution to derisking. In most cases, 80% to 90% of what you want to get in your pension de-risking program can be done in the cash bond markets.

Now, if you're using derivatives because you don't have the ability to get the exposure or hedge with cash investment, derivatives can be another way to utilize capital efficiency within the fixed-income portfolio. To me, that is a risk-reducing exercise. It should not have an impact on your return. We run across clients that do more sophisticated strategies such as swaptions and put spread collars. But those strategies are not right for everyone, and take a tremendous amount of education and the appropriate governance structure.

JAMES SO: Passive investing is a big deal in large-cap equities but less so in fixed income, and much less so in the LDI world. We see some clients with passive or index strategies, but most understand that they either want or need the alpha that we and other LDI managers generate. They either want us to generate alpha to help them cover future service costs, or to keep up with the liabilities, which is pretty hard to do.

Credit downgrades affect the liabilities of the assets, and if you're a passive investor, you would be riding any downgrade all the way down. So as a company's fundamentals are deteriorating, the yields are getting higher and spreads are widening. Once it's officially downgraded by the rating agencies, the bond is punted out of the index and a passive investor, along with all the other passive investors, will have to sell it at the end of the month. They're all rushing for the exits at the same time. So you're getting hurt on the asset side and on the liability side; when the downgrade occurs, your liabilities go up in value.

If you're a passive investor you will bleed funded status over time.

But if I spend five minutes talking about this with a plan sponsor, they get it. Consultants get it, we get it, our competitors all get it, and there's a handful of us that have a really long track record of beating the indexes.

GARY VEERMAN: James is right. You're subject to downgrade risk in your portfolio while benefit payments are effectively risk-free. You're a forced seller of a bond that gets downgraded. A statistic that's so powerful is, if you were 100% funded in 1991, 100% fixed-income duration and credit-quality matched, you would be 84% funded today just through the impact of being passive and not being able to avoid any of those downgraded bonds or adding value through active management. Talk about a strategy that doesn't work.

On the Treasuries side, there are some very cost-sensitive plan sponsors going passive. I think from that side of the equation, it's more of a preference.

JAMES SO: One I would call 'majoring in the minor.' Paying too much attention trying to over-engineer the fixed-income solution. Take care of the big risks first. The big risks are really asset allocation decisions, but if you have an asset allocation that's heavy on return-seeking assets, your proverbial golden ratio — 70% equities, 30% fixed income — the equity, or the return-seeking portion is going to throw off so much risk vs. your liabilities that it doesn't make sense to over-engineer your fixed-income solution. Yet plan sponsors sometimes want to engineer the fixed-income benchmark to squeeze out another 100 or so basis points of tracking error.

The other one is what I could call 'cognitive bias.'

We have clients that want a higher-quality fixed-income portfolio. We can oblige, but we will point out that they're going to miss out on some alpha. The additional risk we might take from being able to go down the credit quality spectrum to the triple-B space or perhaps even cross-over credits could add quite a bit of alpha and not that much tracking error. The tracking error of the portfolio might go from 100 basis points to 150, or 200 basis points — and that concerns them while they're willing to keep a heavy return-seeking allocation that generates 2,000-plus basis points of tracking error vs. their liability.

FRANÇOIS PELLERIN: Some old habits die hard. Many plan sponsors are not hoping that interest rates will go up, they know they will. That's a mistake. We continue to think that even when rates are low, it's a non-compensated source of risk.

For a typical plan, 90% of the risk exposure on rates is beyond the 10-year key duration node, and the Fed doesn't control that. The Fed controls the front end. For both fundamental and technical reasons, we don't see how the long end can go up dramatically and salvage pension plans. I would continue to try to take risk that's compensated.

The other mistake that I see is plan sponsors feeling good about selling their retirees to an insurance company via a partial annuitization. If half of your liability pertains to retirees and you sell all those to an insurance company, you haven't gotten rid of half the risk. You've only gotten rid of a small portion of the risk because you will be mostly left with participants whom many insurance companies don't want to touch with a 10-foot pole.

GARY VEERMAN: I agree with François on the interest rate piece, but I'll say it in a different way. I don't think plans hedge enough interest rate risk. The fact is, if you're not getting paid for it, why are you taking risk? It's as simple as that.

And on the other part of risk, I think in a lot of cases, very well-funded plans, or plans near 100% funded, could be taking too much risk. Obviously, it's hard to generalize the correct amount of risk because every plan is different. But it's hard to argue that increasing funded status beyond a certain point, say 115%, adds any real value to most plans. We refer to it as pension asymmetry, and it can be simply defined as the fact that an increasing funded position has decreasing marginal benefit, while decreases in funded status are increasingly painful.

We think pension asymmetry should factor in a plan's strategic asset allocation decisions.

FRANÇOIS PELLERIN: The average plan now is 80% to 85% funded. People make the mistake of thinking, “Well if interest rates go 150 basis points higher and the equity markets do well, I'm going to magically be fully funded.” But you can't think about that on a snapshot basis, you need to forecast that scenario over three or five years. You then realize that an 85%-funded plan with a bit of service cost won't reach full funding under any reasonable capital market scenarios.

A harsh reality that often has to be digested by plan sponsors is that adopting a sound contribution policy is critical and must work hand in hand with the investment policy. It's about how you have these two components work together so that you get the output you want within a reasonable period of time under a reasonable amount on risk.

GARY VEERMAN: Is there going to be a silver-bullet solution? No, I think it's the opposite. There are plenty of plan sponsors that still have core bonds, and the next evolution for them is moving to long.

I do believe we'll continue to see new entrants into the space and what I would just caution is the more sophisticated the strategy sounds, the more I would question the validity of it.

There have been a lot of asset flows into the space, and I think there are going to be some challenges for managers who start to have books of business in the credit space that are $30, $40, $50 billion. Particularly the way they add returns. Bottom-up security selection works well up to a certain point; ability to be nimble and take advantage of smaller-sized deals, and have those deals impact the portfolio's outcomes — all those things are a reality important for plan sponsors who want their LDI managers to continue to add value consistently.

JAMES SO: From an implementation perspective, the trend of sell equities and buy fixed income — typically long-duration fixed income — will continue. And in the United States, we're very fortunate because we have a huge supply of long bonds.

In other markets around the world, particularly in Europe, they don't have a supply of long bonds. So for them to implement LDI, they match off against the duration beta of their liabilities by using derivatives, generally swaps.

We can do that here as well, but we have a more liquid futures market. It's not necessary because we have such a supply of long bonds. Over the past few years, I've heard the concern that when every corporate defined benefit plan moves into long bonds, there won't be enough supply. My response is that one, we're moving very slowly, and two, there is more than one way of achieving that goal.