Sunday, April 10, 2011

Apples and Oranges?

Following my comment on Ontario Teachers' 2010 results, a reporter contacted me to discuss their results. He was particularly interested in understanding the risks Teachers' was taking to deliver these exceptional results. Teachers' delivered 14.3% in 2010, surpassing the Caisse which gained 13.6% in 2010. My answer to this reporter is that we simply do not have enough public information to understand all the risks that are being taken internally at Teachers', the Caisse or the other large Canadian defined benefit (DB) plans, so comparing results is next to impossible because what ultimately matters is comparing risk-adjusted returns, not just headline figures. But I added: "You'll find out who is taking the most risk the next time markets tank. When you're managing pensions, it's protecting downside risk that matters most and unfortunately that risk is hard to ascertain until a crisis hits. Only then will you see who was taking the biggest risk."

Having said this, there are things we do know. Teachers' is a pension plan which manages assets and liabilities. This means they make decisions on the fund's alpha and beta (policy) portfolio, measure liabilities, and administer the plan. The same goes with OMERS which earned 12% in 2010. Importantly, these funds assume the risk of their policy (beta) portfolio. In contrast, the Caisse, CPPIB, PSP Investments, AIMCo, bcIMC, are fund managers which manage assets. They do not engage explicitly in asset liability (ALM) management like Teachers' and they do not administer their plans. Moreover, they recommend an asset mix to their depositors but they do not assume the risk of the policy (beta) portfolio. For example, at the Caisse, depositors will choose the asset mix that best fits their members' liabilities (they use consultants to help them decide). Once that asset mix is chosen, fund managers can play with the target ranges a little, but the beta risk is assumed by the depositors.

Why is this information important? For two reasons. One, if you're adopting an ALM approach, you're looking for absolute returns with minimal risk. Alternatively, funds managing assets relative to a beta benchmark are looking to add value over a benchmark. Where things get messy is that while everyone pretty much agrees on some standard benchmarks for public markets, there is a lot of variance in private markets. Some use absolute return benchmarks while others use spreads over comparable public market investments.

Second, getting the benchmarks right is critical because that's how pension fund managers measure the value added and that's how their total compensation is determined (here in Canada, typically over a 4-year rolling return period). If the benchmark does not reflect the risks (beta, leverage, illiquidity, etc.) of the underlying investments, then there is a potential for abuse and all that value added can blow up quickly when a bomb explodes.

I mention this because I clearly remember a lunch I had with the Henri-Paul Rousseau, the former President and CEO of the Caisse, in the winter of 2006. He was gracious enough to meet me at Molivos restaurant, one of my favorite Greek restaurants in Montreal. We enjoyed the food and then discussed politics and investments. I remember we talked about benchmarks and risk. I told him that the Caisse has some of the toughest benchmarks to beat in real estate but I also told him I was concerned with illiquidity risk in private and public markets. I will never forget what he told me: "Leo, the way we manage risk, I doubt anyone comes close. We have the highest risk-adjusted returns in the industry."

I was stunned. To be fair, this was a year and a half before the 2007-2008 financial debacle which led to the Caisse's $40 billion train wreck, and nobody was even thinking that markets will blow up in a historic way. In the past, I've been tough on Henri-Paul Rousseau and so has the media, but no matter what you think of him after the debacle, I can tell you this, the man is extremely intelligent. Looking back, I think he had way too much confidence in Richard Guay, the head of risk who then became the CIO under his watch. I praised Richard in the past but it's clear to me now that either he didn't understand the enormous illiquidity risk of non-bank asset backed commercial paper (ABCP) or what was backing this paper, or he willfully closed his eyes as they bought huge chunks of ABCP to handily beat the money market benchmark of T-bills. (And as Diane Urquhart noted, the Caisse wasn't alone in taking big risks in ABCP and other credit derivatives prior to 2007).

The crucial thing to understand here is someone, somewhere wasn't asking the tough questions and totally neglected their fiduciary duty to protect assets. I also question what kind of advice the Caisse's depositors were getting from their consultants. Had they no clue what kind of exposure the Caisse had to ABCP and what was backing these assets?

Following this debacle, Michael Sabia came in to clean up the then 'basket Caisse'. In fairly short order, he focused on risk and putting the right people in key positions to make sure this never happens again. Of course, markets have been rising since Sabia took the helm, which always helps, but beyond that, he's implemented fundamental governance principles based on transparency and accountability and he hasn't hidden the fact that more needs to get done.

At the end of my last comment on the Caisse's 2010 performance, there is a CBC interview with Mr. Sabia in French. At one point the reporter is almost attacking Mr. Sabia on whether or not he's going to give senior managers and employees bonuses. I almost gagged. Mr. Sabia was cool and collective and answered the question properly, stating that the board will determine bonuses based on the fund's and individuals' performance. Had that reporter interviewed me, I would have set him straight and told him that Caisse was the only fund in Canada that didn't pay bonuses after 2008 and they got some of the toughest benchmarks to beat in all their investment portfolios.

Now, let me get back to the topic. If you look at the headline number, Teachers' delivered 14.3% in 2010 whereas the Caisse returned 13.6%. Teachers' added 450 basis points of "value added" over its policy portfolio (benchmark) whereas the Caisse added 410 basis points above its policy portfolio. Does this mean Teachers' results were better than the Caisse? Does it mean the guys and gals at Teachers deserve to get paid a lot more money than the people at the Caisse?

Of course not. The problem with looking at headline figures is that they tell us nothing about what risks the pension fund took to deliver these results. Take the example of a hedge fund which is selling premium all day long. The manager is going to boast that his fund has a "high Sharpe ratio" and while that may be true, you know he's going to get clobbered the next time markets go south. This is a perfect example of why you shouldn't just look at returns and Sharpe ratios.

I vividly remember a senior pension fund manager at a large pension fund which was using the balance sheet of the fund to engage in illiquid 20 and 30 year swaps. Nobody really understood the risks he was taking but as long as he was making money, he was untouchable. I remember arranging a meeting with that manager and two sharp fund of funds managers I know in Montreal who manage money for ultra high net worth individuals. They listened carefully to his strategy. He was prodding them to see if they would fund him. Then one of them told him flat out: "Stay where you are. Nobody in their right mind would be crazy enough to fund this strategy except for a large pension fund with deep pockets and a large balance sheet." They were right. He blew up in 2007-2008 losing billions.

All this to say that performance means nothing to me. I don't care if the manager has a PhD in nuclear physics from MIT, if it looks too good to be true, I would run away fast. In the same vein, saying that Teachers' had a better performance than the Caisse in 2010 doesn't tell me much because I don't know the underlying risks both funds were taking to achieve those returns. On a risk-adjusted basis, the Caisse's results might be a lot better than those of Teachers'.

Then there is the question of value added. A long time ago I discussed bogus benchmarks in alternative investments and followed up with a comment on why we can't properly compare pension funds. Benchmarks matter because that's how you measure "value added" and that's how you determine compensation. Take the example of real estate. The Caisse's Real Estate portfolio achieved a 13.4% return in 2010, outperforming its benchmark index by 1.8%. Teachers' Real Estate portfolio returned 16.9% in 2010, outperforming its benchmark by a whopping 9.2%! Does this mean that Teachers' Real Estate team is better than that of the Caisse's? No way! The Caisse has one of the best Real Estate teams among all the public pension funds in the world.

But it's obvious the Caisse's Real Estate benchmark is more representative of the underlying risks of the Caisse's RE portfolio than that of Teachers' benchmark for RE (CPI + 500 bps). I don't care if it's hedge funds, private equity, real estate, infrastructure, or even money markets, when you see a pension fund manager trouncing his or her benchmark, then you'd better be asking some tough questions. Teachers' will claim that they do asset liability matching (ALM) and therefore need to have absolute return benchmarks for every investment portfolio but I question the logic of having benchmarks that don't reflect the risks of the underlying portfolio (read more on benchmarking real estate by clicking on this Townsend Group study).

On internal alpha strategies, pension funds should clearly report the benchmarks they're using as well as the notional and economic exposure of their derivatives strategies. For example, using a lot of leverage in fixed income relative value strategy is fine for short term bonds, not so for long term bonds. Also, if a pension fund is engaging in direct real estate and private equity deals, they should discuss the performance of this activity (since inception) apart from that of fund investments.

Importantly, pension funds should disclose the benchmarks for each and every investment activity and explain why they're using these benchmarks. They should also track historic changes to benchmarks with reasons as to why changes were made and make such a document publicly available. I will be the first applaud the first Canadian fund that has the guts to publish such a document.

Benchmarks aside, one senior pension fund manager wrote me to tell me what should be reported for all asset classes is cumulative IRR since a strategy's inception: "If we are long term investors we should act like it. Average annual returns, and mutual fund like 3 and 5 year average data, etc. does not capture the fact that most institutions scale up at wrong times."

What are the other risks that are important when comparing large pension funds? I would place operational and culture risk right up there. Operational risk covers operations, systems processes, and the risk of fraud, which I referred to recently in the CalPERS's case. What do I mean by "culture risk"? This is typically linked to key person risk. For example, is your Head of Legal an incompetent jerk? Did you properly vet him or her before hiring them? Is your Head of Real Estate an egomaniac who shovels billions to his favorite fund in hopes of setting himself up on Easy Street later on? If you don't vet your leaders properly, you're creating a poisonous culture.

More importantly, are your analysts, back office and middle office staff happy and engaged in the process or do they feel totally discouraged and demoralized? On the flip side, are your employees grossly arrogant when they call brokers and other external partners? (I'm hearing horror stories out there about some of the large Canadian pension funds). Culture is the single most important thing in any organization and yet few leaders take the time to address this critical issue. (Make sure your employees are respecting Ray Dalio's Principle #11. If they don't, they're just slimy weasels.)

Let me conclude by stating there is no perfect pension fund. There are good and bad people working at all the major pension funds in Canada. This business is full of egos and some of the biggest egos are all the way at the top (I'll never forget Clause Lamoureux telling me to "shut up" after I testified at the pension hearings in Ottawa). So many fragile egos and yet I respect people who can fess up and admit they made a mistake. In this business, take your ego and shove it!

No comments:

Post a Comment