CPPIB dynamically reviews its total portfolio

The CPPIB is considering the next phase in its total portfolio approach to managing assets, allowing for a more dynamic funding of investments from the policy portfolio, as the nature of the assets in the real portfolio change.

The total portfolio approach adopted by the Canada Pension Plan investment Board (CPPIB), is a unique structure which dictates that any move away from the policy portfolio, to the real portfolio must be funded by a corresponding shift in assets in that policy portfolio.

As with other funds, the CPPIB has a reference portfolio, which represents the low cost, low complexity investment strategy; and then its real portfolio is an active decision away from that reference. What makes it different is in the funding.

“We make investment decisions, then ‘ask how do I fund it?’ There are two parts,” chief investment officer, Mark Wiseman says.

“Most asset managers are good with the first decision. We have a total portfolio management group which are essentially responsible for making the funding decision.”

The fund is now looking at the next step in this model, which is being more dynamic in the funding, he told delegates at the recent CFA Institute asset and risk allocation conference.

“We could look at changing the funding status as the nature of the underlying investment changes, for example, as it delivers. That is the next phase of the total portfolio management approach, but then we have to look at who is responsible for that.”

For the first five years of the CPPIB, the low cost, low complexity strategy was followed with a passive allocation split 65 per cent in equities (15 per cent in Canada, 45 per cent international equities, and 5 per cent emerging markets), and 35 per cent in bonds (25 per cent Canadian nominal, 5 per cent real return bonds and 5 per cent foreign sovereign).

Wiseman says the low level complexity of this approach means it could be done with a dozen people in the office, but it does provide a good benchmark.

“This is a very real benchmark, it represents the path ‘not taken’, and we can measure the exact volatility and value added,” he told delegates.

The new total portfolio approach was first implemented six years ago, where active decisions are taken beyond the reference portfolio which has the risk tolerance policy setters are comfortable with. Any move away from the reference portfolio has a cost associated with it.

“The third part is to manage risk exposures and not the asset class labels,” Wiseman says. “We want to control the risks not the notional value, and charge the investment program an opportunity cost. We manage risks such as liquidity and currency at the total portfolio level.”

This approach means there are no allocation limits on asset classes, rather it is about risk/reward tradeoffs.

For example, the reference portfolio’s entire equity allocation of 65 per cent could be allocated to private equity.

At the moment the total active risk is about 200 basis points – with real estate and public equity about 50 basis points each, and private assets about 100 basis points.

“Everything we do is an active overlay over our reference portfolio,” Wiseman says. “We will make a series of small, very well considered actions as opposed to macro tilts on the total portfolio because the amount of active risk they make are extremely large compared to smaller decisions. For example buying 35 per cent equities and selling 35 per cent fixed income, creates 600 basis points of additional active risk.”

When contributions come in they are immediately invested in the reference portfolio. Any decision to allocate investments outside of that reference portfolio has to be countered with a funding decision from the reference portfolio which reflects the specific risk/return attributes of that particular investment rather than assuming homogeneity within an asset class.

For example for a private equity investment of $100 million in a US tech company, Wiseman says the team would look at funding that minimises the active risk, so public equity with a similar characteristic, and in this case, selling a sub-index of technology stocks.

The investment team at CPPIB divides the world into 70 cells, 10 industry sectors and seven regions, and looks at the funding out of its reference portfolio to match the economic exposure in the investment it “buys” in the real portfolio. It is more complex that traditional asset allocation and requires complementary technology, risk management, analysis and decision-making frameworks to make it effective.

Wiseman says in this example the average private equity beta is 1.3, so the passive portfolio is debited of S&P tech index stocks of $130 million, and the active portfolio is credited with the new investment of $100 million in private equity and $30 million of debt in the reference portfolio.

“We literally do it on an asset by asset basis, depending on the nature of the underlying investments, for example an airport in India, we fund it with emerging market equities sales, it entails getting below the asset labels,” Wiseman says. “At the end of every year we can look at what the reference portfolio would have produced and the return of the total fund between the two and measure our active risk and added value.”

The investments of the CPPIB are essentially a series of long and short decisions, that take it from the reference to the real portfolio, he says.

Each investment is assessed on its own merit, for example in real estate, the funding of the investment will be very different for a core office building compared to a condo in Palm Beach because of the underlying nature.

This next step is to change the nature of the funding dynamically, as the beta changes.

“For example if a real estate manager levers the building in Manhattan we’ll charge our portfolio more, so sell equities. Our investment professionals have tended to be more conservative investments, so our real estate portfolio is largely unlevered. There is no incentive for them to become more risky – they don’t get paid for it.”

The CPPIB is in a unique situation for many reasons – it doesn’t have to pay anything out for 11 years and in the next 27 years doesn’t have to pay more than 4 per cent a year – but also in the amount of risk it can take. As a result it has been investment entrepreneurial.

“Our problem is how do we find enough good investments out there? How do we create enough active risk to meaningfully outperform the reference portfolio. The board says the risk limit is 450 basis points, and we have to look at how to get there without just making a big macro tilt, so we are extremely open to investment ideas that contribute to risk,” he says.

The reference portfolio is reviewed every two years: foreign sovereign bonds were most recently added and the Canadian equities exposure is shrinking, but the 65:35 hasn’t changed.