Vintage year funds launched in a weak market offer the best returns, but spread investments across multiple vintages, says Jeremy Plummer

In the private equity world the importance of diversification by vintage year is well understood. Returns vary considerably between vintage year (year of inception) of private equity funds, and picking the right vintage can be more important than picking the right fund within a vintage. This is simply a product of market pricing cyclicality: a vintage year when corporate valuations, hence acquisition prices, are high will have lower returns than a vintage year when they are low.

It can be difficult to recognise the top and bottom of the corporate valuation cycle, so it makes sense to diversify investment across a range of different vintage years, a mix that should deliver attractive returns over the whole business cycle.

Vintage year is similarly a key determinant of returns in private real estate funds. It is already clear that the 2006 and 2007 vintages, established when real estate valuations were at high cyclical levels, will produce very disappointing returns. By contrast, vintages beginning in years when market sentiment was weak and valuations were low, have produced very good returns. Perhaps not surprisingly there is evidence of an inverse correlation between the strength of the economy at the time of inception and the returns of a vintage. Vintages starting during or after a recession, when property values have fallen, are likely to be best.

Figure 1 shows the average returns of different vintages of real estate funds, starting between 1996 and 2007. The best vintage, on average, was 2001, which coincided with a weak period in property prices following the collapse of the tech boom.

On this basis, it is likely that the 2010 vintage of real estate funds will be a very good one. Nevertheless, before investing all one's available capital into this vintage, it is important to remember that vintage year diversification is not just about timing of market entry. Another key reason to diversify by vintage year is to spread the duration profile of fund investments, in order to avoid a situation where all funds in a portfolio mature at the same time, resulting in significant re-investment risk.

A well-constructed portfolio of real estate funds should aim to have a broad spread of maturity dates, thereby minimising re-investment risk. A well spread maturity profile will also have the added benefit of providing a regular source of natural liquidity in the portfolio, offering the opportunity to re-balance the portfolio each year into preferred strategies for a particular vintage.

Constructing a portfolio of funds with a good spread of maturity dates requires patience at the outset. Given that the duration, or average hold period of assets in private real estate funds ranges from three to 10 years, the time required to build a portfolio with full vintage year diversification is at least five years. It is possible to accelerate the construction process and gain quicker market exposure by acquiring secondary interests in existing funds, which are due to mature over a shorter timescale.

Figure 2 illustrates the cash flow profile of a portfolio of funds assembled over four years, with a front-end loaded market entry using secondaries. Some of the secondaries start to mature within four years and the new funds have a mix of durations ranging from three to eight years. Commitments are made over the first four years and the portfolio reaches ‘maturity' from year six onwards. Maturity is characterised by a consistent level of annual realisations, representing around 15% pa of the total portfolio. These realisation proceeds are assumed to be re-invested in the chart below.

A further point highlighted in the cash flow graph is that the portfolio always has commitments to underlying funds which are not yet drawn. Investors are therefore faced with two choices: either they can accept that a proportion of their total commitments will be held in cash, waiting to be called, or in order to be fully invested in cash terms it is necessary to ‘over-commit', ie, to commit more capital than was intended.

The amount of over-commitment required to be fully cash invested depends on the nature of underlying fund investments. The higher the allocation to funds with long investment periods, the greater the over-commitment required. Over-commitment carries obvious risks of being called for more cash than intended in times when markets are illiquid. This has become a serious issue recently for over-committed investors, who have suffered serious liquidity problems during the market crisis. In our portfolios we are cautious not to over-commit, but limit the amount of inefficient cash holdings by favouring investments with immediate capital call, or relatively short investment periods.

In the boom years of 2006-07 a number of funds of funds were launched with open-ended, or semi-open-ended structures. These structures have attracted criticism for investing in closed-ended real estate funds, which are fundamentally illiquid, so that there is realistically very little prospect of the fund of funds being able to fund redemptions without making sales of illiquid assets at discounts, which is not in the interest of the remaining investors in the fund.

The error these funds have made is the failure to match the expected liquidity of the underlying portfolio with the promised liquidity at the fund of funds level. The solution to this is to construct the portfolio so that it does provide a match. In our global funds of funds, we construct portfolios from the outset with a mix of durations, so that the natural liquidity in the portfolio is matched with the promised levels of liquidity over time.

We believe that the 2010 vintage of real estate funds will be a great one, which will provide exceptionally high returns. However, investors should pause before investing all their money in a single vintage. It makes sense to front load the timing of market entry to capture the low point of the market cycle. Investors with a long-term horizon should take the time to construct a portfolio with a well spread maturity profile, as this will pay dividends over the long term.