For any managers hoping to appeal to the California State Teachers’ Retirement System, the best route is to go direct. Like most of its peer group, the second-largest institutional investor in the US historically leaned on external managers, specifically via their commingled funds, for exposure to private markets. However, to exercise more control over its portfolio, cut fees and, ultimately, achieve better returns, the $240 billion pension is making a concerted effort to manage more assets internally.
Although the system has participated in separately managed accounts, co-investments and joint ventures for decades, particularly through its real estate allocation, it has adopted a new portfolio-wide strategy that brings these alternative structures to the forefront. This ‘collaborative model’ calls for CalSTRS to play a more active role in its investments, be it through bespoke club deals, partnerships with other institutional investors, or by acquiring operating companies to act on its behalf.
The collaborative model is equal parts investment thesis and branding tool. CalSTRS believes it can achieve substantial cost savings by keeping assets in-house. Justifiably so – in 2017, it managed 44 percent of its portfolio internally at a cost of $30 million compared with the $1.8 billion it paid external firms in management fees for the other 56 percent. It also wants to send a signal to managers large and small, as well as to other investors, that it is open for business, so long as the arrangement fits its terms.
“This is a vision that we want to execute across the asset classes,” says CalSTRS deputy chief investment officer Scott Chan. “It’s become the most meaningful implementation platform for CalSTRS and we want to be able to communicate it well to the market, to our potential partners, to our peers, to the [state] legislature, to the [CalSTRS] board and to our clients.”
Led by CIO Christopher Ailman, the CalSTRS investment team identified the collaborative model as a system-wide priority in 2017. The decision came after an internal evaluation found that 97 percent of its non-carried interest expenses were going to outside managers. The following year, Chan was hired away from the University of California Regents where he headed the governing board’s $55 billion global equities portfolio. As second in command at CalSTRS, it is his task to implement the vision.
This past August, Chan and Mike DiRé, director of real estate, sat down with sister title PERE at its Sacramento headquarters to discuss the collaborative model, the implementation process and what it means for the pension.
One strategy, many styles
All asset classes, both public and private, will fall under the umbrella of the collaborative model, but CalSTRS will approach each allocation slightly differently. For real estate, it will focus on acquiring operating companies and forming joint ventures with sector specialists. To facilitate more co-investment in private equity, it will empower its staff to make swifter commitments. As it builds a more customised infrastructure portfolio, it plans to seek out like-minded institutions to invest alongside.
“It’s not a one-size-fits-all model,” Chan says. “It’d be wrong for us to have a one-size-fits-all strategy where we’d force asset classes that may not be ready to execute different strategies into that mould.”
One of CalSTRS’ biggest gripes with closed-ended fund structures is their rigidity. As a limited partner, it has no meaningful say about what assets are acquired, or when they are bought and sold. With flexibility factoring so heavily into the collaborative model, Chan says he is reluctant to weigh it down with such mandates and hard targets.
On 5 September, CalSTRS’ investment committee rolled out its new strategic asset allocation plan. It calls for a 2 percent increase to both its real assets portfolios: real estate and inflation-sensitive, the latter consisting of infrastructure and inflation-linked securities. Targeting 15 percent and 6 percent, respectively, the committee hopes to get a premium from the two illiquid asset classes. It will offset these increases by shaving 5 percent off its public equities exposure. It will also add 1 percent to its risk mitigating portfolio. Although the collaborative model was considered during the drafting of this new strategy, it was not factored into the expected results. Its annual return target will hold steady at 7 percent.
“We have not included the benefits of better implementation or active management in determining our new strategic asset allocation because we want to be conservative in the return/risk forecasts,” Chan says. “Instead, we’re forecasting ‘beta’ returns, risk and correlations on a very long-term basis and applying the collective wisdom and judgement of the team to come up with a strategic asset allocation.”
The real estate team hopes to hold 70 percent of its assets in long-term structures; private equity will try to double its co-investment exposure from 7.5 percent of its allocation to 15 percent – which would equate to a jump from $1.6 billion today to $3.2 billion – and, overall, Chan hopes CalSTRS will save between $300 million and $500 million over the next five years. Otherwise, unlike the strict, visible framework of the fund’s strategic asset allocation and its 500-plus benchmarks, this philosophy will largely play out behind the scenes.
“This isn’t a model that is closed,” DiRé says. “It’s not like we’re getting rid of all our manager relationships, or we’re only open to a certain structure going forward. CalSTRS is open for business and we just want to take a more progressive thought toward the way we structure relationships going forward across asset classes.”
CalSTRS’ real estate team has set the standard for the collaborative model. It launched its debut separately managed account in 1987 and joint venture in 2002, according to May meeting documents. In 2007, it purchased its first operating company. As of 31 March, 95 percent of the fund’s core real estate was held in what it considers active structures and it has consistently beaten its benchmark, the NCREIF Open-end Diversified Core Equity index. CalSTRS’ core portfolio has achieved one-year net returns of 8.47 percent, three-year net returns of 8.72 percent and five-year net returns of 10.39 percent, compared with 6.55 percent, 7.01 percent and 9.18 percent, respectively, for ODCE.
Overall, CalSTRS’ top 15 real estate managers, which account for 81 percent of its net asset value, have an estimated weighted average IRR of 7.7 percent since inception, according to meeting documents, compared with the 5.7 percent produced by the portfolio at large.
Most collaborative model investments will focus on lower risk strategies, Chan says. However, having a closer alignment with its managers will allow CalSTRS to “create a better and more intentional risk culture” by more easily comparing investments across asset classes.
“In the margin, we can understand the risks we’re taking against our strategic asset allocation, what return we’re getting and if they are commensurate with what we expect, or if we should be doing something else,” he says. “It allows you to create a common language to compare across assets and have your asset classes compete against each other.”
The ‘partner of choice’
To participate in more direct transactions more frequently, CalSTRS will need to increase its capacity to find, underwrite and execute deals. It plans to beef up its investment team over the next five years, possibly growing the staff from 180 to more than 300. But that would be just the tip of the iceberg, Chan explains.
“For the organisation, it’s a bit like throwing a pebble into a pond and seeing a ripple effect,” he says. “If we’re growing our investment organisation, we’re going to need more legal support, more tech support, more procurement support, et cetera, et cetera. This is an effort where we need to get to a point where we become more and more the partner of choice. And to do that, it’s a whole organisational effort.”
In addition to attracting more talent, CalSTRS is committed to retaining it. As Chan notes, it would be difficult for partners to throw support behind an organisation with frequent turnover. However, this is easier said than done, as investment professionals who rise through the pension’s ranks often draw the attention of private employers capable of paying higher salaries and offering better incentives.
Herein lies a central issue for CalSTRS and other public pensions interested in building robust internal investment teams: they must spend on compensation, travel and other expenses to save money in the long run. In the US, where top talent has plenty of options and constituents are wary of government spending, staff pay has proven to be a difficult hurdle to clear.
Despite broader constraints on staff pay, DiRé says CalSTRS has been well supported in its efforts to buck the trend. “The board has been very progressive over the years to move in the direction of creating salary structures that work.” Continuing to do so is imperative, Chan says, to building a team worthy of the partnerships it seeks and to keep pace in what is an increasingly competitive field. “If we looked at a peer group in APG, GIC, any of the Canadian funds, they have more people, they’re moving faster than we are and they don’t have any of the rules or regulations that we have to work within as a state agency. Their only rule is to make money.”
The CalSTRS compensation committee is reviewing the system’s pay scales too, according to the fund’s meeting documents. It plans to adopt new salary ranges for investment managers in March 2020 then implement them the following July.
Other pension funds that have pursued direct investment models have set up satellite offices to attract talent and get broader exposure to dealflow. The Teachers’ Retirement System of Ohio, which has had a direct investment programme since the 1980s, has outposts in New York, San Francisco and Atlanta. The Teacher Retirement System of Texas, which has a newer direct investment programme, opened a London office in 2015 and has its eyes on a Singaporean location as well. Although CalSTRS has no plans to branch out beyond Sacramento – where it is adding additional office space – Chan said the fund might consider adding secondary locations or opening the door to remote employees.
“We’ve tapped a lot of great talent in this location in Sacramento, but we’ve got to be realistic in thinking about how much talent also resides in San Francisco, LA and other areas,” he says. “We’ve got to be creative in thinking about that, particularly if we think on the investment side that we’ll eventually go from 180 to a little bit over 300.”
A growing trend
Like some other internationally investing institutions, the large US pensions have tilted their focus away from blind pool funds since the global financial crisis. Texas TRS, for instance, rolled out its principal investments programme in 2014 and has already seen tangible results. When it implemented the strategy, just 24 percent of its real estate allocation went toward non-fund structures, compared with 60 percent last year. During that period, its internal real estate holdings achieved a 14.4 percent IRR compared with 10.3 percent secured by its external managers.
Other private asset classes have performed even better under the principal investment model. Texas TRS’s direct investments in energy, natural resources and infrastructure produced a five-year IRR of 14.3 percent compared with 3.6 percent for the equivalent fund portfolio. During a public meeting in July, CIO Jerry Albright attributed the success to a single acquisition – which he did not identify because of the organisation’s non-disclosure rules – that would not have been available through a commingled fund: “A lot of that return came off of one transaction … and we made some fantastic money that we wouldn’t have made had we just accepted what the partner brought to us.”
CalSTRS’ cousin, the California Public Employees’ Retirement System, is also targeting a more advanced approach to direct investment: it aims to launch a pair of investment companies that would operate independently of the pension. This has become known as the ‘Canadian model,’ which refers to retirement systems in Canada that have launched standalone investment arms: Cadillac Fairview by Ontario Teachers’ Pension Plan, Oxford Properties Group by the Ontario Municipal Employees Retirement System, and Ivanhoé Cambridge by Caisse de dépôt et placement du Québec.
Of the decision to go direct, Eric Plesman, Oxford Properties Group’s executive vice-president of North America, says a lot of factors come into play, including an institution’s size and resources. But he says: “It ultimately comes down to the mandate, the level of control you want to have in your investment strategy and how much capital you need to deploy. It depends, but for us, it’s been essential to the returns we’ve realised to date.”
Ben Maslan, managing director of RCLCO, a real estate consultant hired by CalSTRS last year, says more US investors are looking to emulate the Canadian model by better aligning their interests with those of their managers. However, there is a significant amount of daylight between the two approaches, particularly when it comes to launching standalone entities. “The Canadian model is either an operating company itself investing directly in real estate, or it will partner with an operator to invest directly in real estate,” Maslan says. “That latter part is where we see the American model moving, partnering with operators.”
One such partnership is CalSTRS’ majority ownership in Fairfield Residential, a multifamily real estate operating company based in San Diego, California. Earlier this year, CalSTRS bought a 65 percent stake in the firm from Brookfield Asset Management – the two entities recapitalised Fairfield in 2010 and brought it out of bankruptcy.
With roughly 1,200 employees and assets in 36 markets across the US, DiRé says the Fairfield investment is a prime example of how the collaborative model can expand CalSTRS’ reach without overloading its internal capabilities. “If we can structure relationships that work for both parties, we don’t need to hire those people,” he says. “We have those people through the relationships and it’s much more efficient for us. We’re not going to grow to that size of a staff. It just wouldn’t make sense to run it, especially out of Sacramento.”
Another pre-existing structure that fits the collaborative model is an infrastructure club deal between CalSTRS and Dutch pension manager APG, which is managed by Argo Infrastructure Partners, a New York-based manager. In 2015, the two investors chipped in $250 million to the initial venture, which acquired Cross-Sound Cable, a high voltage direct current transmission system between Connecticut and New York. Last year, both committed an additional $300 million.
“This more direct style of investing not just saves fees, but puts us in a position to be more nimble in the marketplace,” DiRé says. “In the case of real estate, having a direct relationship with developers or even asset allocators, [so] that we can move faster to take advantage of opportunities, just gets us better dealflow, size and structures of our relationships where we feel we have alignment of interests. That gets us fee savings, and the combination of those two things means we should yield higher returns.”
Managing the managers
As more large investors opt for direct approaches, managers are faced with a decision: accept reduced fee revenue or look for capital elsewhere. Although some firms are unwilling to adapt to these more collaborative approaches, many have been happy to accommodate. “It’s a little bit bifurcated for managers,” says Walter Stackler, managing partner of Shelter Rock Capital Advisors. “Some fully discretionary fund managers are happy with what they have and don’t want to change to a so-called collaborative, non-discretionary model. But it works well for the smaller managers as well as larger managers looking to expand their business into new product lines.”
Several top managers, including those that have executed non-fund investments with CalSTRS, either declined to comment or were not available.
A managing director of another capital advisory firm who declined to be named says they have seen an influx of work in non-fund structures. Along with the desire to cut fees, many investors, including CalSTRS, want to hold on to stable, income-producing assets longer than most closed-ended fund structures allow.
Others want the ability to say yes or no to certain acquisitions. Some of these arrangements can even be favourable to managers, the managing director says: “The manager may prefer a smaller team at the limited partner level because they don’t have to deal with a larger team of people micromanaging them and poking holes in their assumptions as much.”
Chan says established managers have been more accommodating of the collaborative model than their smaller contemporaries. DiRé says upstart managers hoping to take on discretionary capital would do well to prove their worth with more collaborative approaches: “It’s still managing money. It’s still managing strategies.”