How 'alternative' are alternatives?

You will notice that I persist in putting the word 'alternative' in inverted commas. I apologise if the reader finds this habit irritating but I seek only to emphasise that I find it very difficult to understand how this word can continue to be applied to asset classes which have proliferated around the world (some estimates now put the number of hedge funds worldwide in excess of 10,000), form the core of many portfolios (usually about 65 per cent in the case of Yale), and have in some cases been around for a very long time (property pre-dates quoted equities as an investment by at least 2,000 years).

As for private equity, again estimates vary but it is clear that well over $300bn was raised globally for private equity funds last year,⁴ and that this figure is increasing steadily and rapidly. This year has already seen the birth of the world's first ever $20bn buy-out fund. With institutional allocations increasing all the time, nobody seems willing to predict just what size this asset class may reach. Already, no company in the world is beyond its reach and a number of large companies, including high street names such as Debenhams and Boots, have been successfully stalked by private equity houses.

In terms of size, then, private equity certainly does not justify the 'alternative' tag, particularly when one considers that the effective purchasing power of these funds is very much greater than their official size, since their transactions are leveraged by debt and mezzanine.

As with hedge funds and property, perhaps the perceived illiquidity of the asset class is an issue. Indeed, there seems to be a tendency among trustees and consultants alike to regard 'illiquid' and 'alternative' assets as the same; the former is used as a definition of the latter. Which may be fine, except for three very important points, all of which seem to be totally overlooked by this very simplistic approach.

Three objections to the traditional view of liquidity

It is sadly typical of the unthinking approach so often found in pension funds that nobody should actually stop to ask themselves why the issue of liquidity should be relevant in the first place. As any accountant or analyst knows, liquidity is relevant only when issues of short-term solvency fall to be considered. This cannot possibly be a material consideration for a pension fund, which has long-term liabilities and thus logically should also be adopting a long-term investment strategy. Both solvency and liquidity are irrelevant. There is not a pension fund in the world, no matter how poorly funded, that has a short-term solvency problem. In the failure by regulators in countries such as the Netherlands, Sweden and Denmark to recognise this basic fact lie the seeds of a vast and totally avoidable tragedy which we will see played out in the years to come at the expense of future generations of pensioners.

Secondly, the belief that legally defined liquidity creates the ability to convert securities into cash at a moment's notice is naïve and unsound. Many investors on Black Monday were unable to sell even FTSE100 shares as the demand for dealing capacity simply overwhelmed the system (not to mention, let it be whispered, that many traders stopped answering their phones). Anyone who has tried to sell a large parcel of shares on one of Europe's smaller exchanges will also have found that quoted liquidity can be illusory, at least at anything like the quoted price. As for the corporate bonds with which many pension funds have stuffed their portfolios, these enjoy liquidity in the legal sense but can be anywhere between difficult and impossible to sell at anything like the quoted price since this market is driven by fluctuating investor sentiment. Anybody trying to sell bonds in 1991 after the downgrade of RJR Nabisco, the indictment of Michael Milken and the indictment and bankruptcy of Drexel Burnham Lambert would have found this out for themselves. Nor does apparent liquidity protect you from the risk of default by the issuer.

Thirdly, the other asset classes are not in fact nearly so illiquid as they may at first appear to those who do not fully understand them. There is a thriving secondary market for private equity partnership interests. Property, certainly prime property in a prime location, is always saleable. Hedge funds are nominally quoted and offer liquidity through redemption, albeit often subject to all sorts of qualifications. I do not suggest that any of these can be turned into cash instantly, but then a pension fund would never need to do this. Most can be turned into cash within a 3–6-month period, at least private equity and property anyway, and that would be ample for any possible requirement that a pension fund might experience.

Why liquidity (or lack of it) is important?

What does any of this matter? Quite simply, because pension trustees and consultants have fundamentally misunderstood not just their need for liquidity, but the adverse consequences of having way too much of it.

David Swensen puts it very succinctly in his book. Liquidity, he points out, comes at a very heavy price in the shape of lower returns. Investors have long recognised that there is an illiquidity premium, that is, that illiquid investments offer a higher rate of return to compensate an investor for being prepared to accept their illiquid nature. The reverse of this must therefore also be true. Liquid investments offer a lower rate of return. Thus, purely from an investment return point of view, and assuming that no other considerations were relevant, an investor would only invest in illiquid assets and would never hold any liquid positions at all.

Of course, other considerations are relevant, at least to some degree, and any practical approach must recognise this. Pension funds can face unexpected calls on their resources, for example as a result of a redundancy and/or early retirement programme, although these are few and far between and even here we are not talking about a need for 'next day' cash, but rather for cash over a number of years, and not beginning to impact on the plan for some months. As we have seen above, most 'alternatives' offer sufficient practical liquidity for these purposes already.

More of an issue is the ability to ride out adverse market conditions. Of course, if you have an uncorrelated portfolio in the way that investors such as Yale do then by definition it is extremely unlikely that all of the elements of your portfolio would be heading south at the same time, but European pension funds do not believe this, despite the principle's endorsement by none less than Harry Markowitz,⁵ or else they would have adopted such an approach. Instead, they have chosen the huge risk of having almost all their portfolio in one perfectly correlated bucket. Thus, if they are ever to be persuaded away from such a method, any such attempt must recognise that they will require reassurance against such an eventuality, even though it be imaginary rather than real.

It will therefore probably be appropriate to provide a comfort zone or safety net in the form of holding the next three years' anticipated net outflows at any one time in bonds, the duration of which match the timing of the individual cashflows as closely as possible. For the typical European DB plan this will usually be somewhere around 8 per cent of NAV. The logic behind such a step is that there has not been any instance over the last quarter of a century when any one asset class (with the exception only of Japanese equities) has had more than three successive years of negative returns. Thus by holding a three-year doomsday reserve one should be guaranteed never to have to be a forced seller of any asset into a falling market.

The true nature of bonds

The widespread misunderstanding of the true nature of bonds is another of the real tragedies of the pension world. That trustees should honestly and sincerely be mistaken in their thinking is perhaps at least partly understandable; after all, they have no financial training or investment experience. That consultants should encourage them in such mistaken beliefs is neither understandable nor forgivable.

There are few European pension funds that calculate their target rate of return at all, let alone on the basis of their actual liabilities as opposed to discounted liabilities and unrealistic demographic assumptions. Instead, they simply pluck an asset mix out of thin air and whatever figure this is likely to produce becomes their target rate of return. The absurdity of such an approach hardly needs pointing out, except that nobody else does, so I might as well. I have discussed this at length in previous issues of this Journal.

My point here is that if pension trustees did indeed carry out such an exercise they would probably discover that their target return was anywhere between about 9 and 12 per cent based on their individual liability profiles and levels of funding. Is there anyone who believes bonds can produce such levels of return? Even at its most conservative, such a target rate would require a real return of at least 4 per cent (and probably significantly more) above the rate of inflation. Yet over the last 50 years UK bond have produced a real return of only about 1 per cent. Thus, holding any bonds at all within a pension plan portfolio can only possibly act as a drag on the overall investment performance, meaning that the remaining assets have to struggle to achieve unrealistically high returns themselves to try to make good the difference. As I have said elsewhere, having bonds in your portfolio is like trying to drive your car with the handbrake on.

Risk

'Ah', the pension consultant will say, 'but you are ignoring risk. Bonds are a very low risk investment, and thus give an acceptable risk adjusted return'.

In fact, none of this is really true. Bonds are not necessarily a safe investment at all, particularly once you take inflation into account. If you had invested in UK bonds at the beginning of the 1980s you would have to wait nearly a decade not to make a positive return, but simply to regain the real value of your initial capital. If you had been a taxpayer rather than a tax exempt investor, then the situation would have been even worse. At the same time, talk of 'risk' and 'risk-adjusted returns' fails to point out that what is being discussed is not really 'risk' at all (whatever this may be, since arguably it changes from one situation to another, and will probably also be different for each individual investor), but 'volatility'.

Once you make the sloppy mental substitution of 'volatility' for 'risk' then you are off into the whole parallel universe which most of the European pension industry inhabit. The one advantage which a pension fund has over other investors is a very long investment time horizon. Thus, just as liquidity is irrelevant, so is volatility. Provided that an asset offers high compound returns over a long period, then the volatility of its returns is immaterial. If, say, UK property offers a long-term compound return of about 10.5 per cent, then does it really matter if its annual returns over the last 25 years have ranged between - 8 and +28 per cent? Provided that the pension plan's target rate of investment return is below 10.5 per cent, then a decision to invest in this asset class would be a sound one. Volatility is irrelevant.

In fact, one could argue that the only material risk with which pension trustees need be concerned is the risk of not achieving their target rate of return, expressed as a compound return over time. With bonds, that risk is likely to be 100 per cent; in fact, there is no uncertainty whatsoever as to the likely outcome. With something at the other end of the spectrum, such as private equity, the return risk is likely to be very low. Thus, the traditional view of bonds as 'low risk' and private equity as 'high risk' is arguably totally the wrong way round. In fact, as pointed out above, bonds have such high levels of return risk that they could not normally be accommodated within a pension fund portfolio at all.

In summary, we have somebody acknowledged as one of the world's great investment thinkers (Swensen) advocating one approach and pension trustees across Europe doing exactly the opposite, apparently on the advice of their consultants that this is the correct 'risk-adjusted' route to follow. Clearly something is fundamentally wrong with this picture, which raises some very important questions, such as 'why haven't pension consultants been telling trustees about the Yale Model?', 'on what grounds do consultants think they are right and Swensen is wrong?' and 'why do they persist in advising trustees to make asset allocation decisions which make no logical sense given their true liability profile?'

Assets and liabilities

To understand why trustees and consultants have such a profoundly mistaken view of risk and return, it is necessary briefly to examine the way in which they view liabilities. I have written about this at length in former issues of the Journal and have no wish to repeat myself. Suffice it to say that there is a very basic misunderstanding at work here.

Because IAS19 (FRS17 in the UK) provides a consistent accounting standard for the treatment of pension scheme liabilities it is quite properly and validly used to calculate a figure to insert in the sponsor's balance sheet. The misunderstanding arises when trustees adopt this figure unquestioningly as the true measure of their own future liabilities to their members. In fact, as will I hope be obvious, it is not this at all and nor was it ever intended to be taken as such. It is an accounting measure designed by the audit profession to provide a consistent approach to preparing a company's financial accounts, not a starting point for investment strategy.

It cannot be a true measure of pension scheme liabilities for three main reasons. First, the assumptions used in assessing the future cashflows are very artificial, particularly when it comes to longevity assumptions (and, despite the stated function of IAS19 to bring consistency of treatment, these sometimes differ from one scheme to another). Secondly, the cashflows thus arrived at are then discounted at a discount rate which is arbitrarily set as the bond rate, whereas what trustees need to consider are the actual cashflows as and when they are likely to occur, not an artificially created present value. Thirdly, longevity assumptions are changing rapidly and unpredictably.

It is this premise that lends credence to the idea of pension schemes investing in bonds. If the 'liabilities' are calculated by reference to the bond rate, then surely by investing in bonds one can match one's liabilities? Should the scheme be in IAS19 deficit then, the argument runs, this still holds true since the trustees can simply demand that the sponsor makes good the current deficit. If necessary, as I heard one consultant say publicly at an LDI conference, 'that's what the PPF is for'.

In practice, this is a tragically mistaken view. Even if the 'liabilities' under consideration were the true liabilities (which they are not), then there are variables at work, such as interest rates, inflation and longevity which make it very difficult just to lock the controls and stand back. LDI vendors would argue that it is possible to insulate a scheme from all of these. I would agree to a limited extent. It is possible to seek protection against the first two, although each extra layer of protection comes at a significant price and even then I am not fully convinced about how well these would work in the event of a large non-parallel shift in the yield curve. As to the third, it would seem logical for a market in longevity risk to grow up around the life insurance industry, but I have yet to see a product on offer that would offer true protection against the full possible extent of the problem. Recent articles suggest that over a half of all 30-year olds in the UK will live to at least a hundred.

Faced with a promise that their liabilities can be fully covered, trustees are understandably confused, and at the same time made nervous of any form of investment other than bonds. Indeed, this is exactly what LDI vendors (who are often simply bond salesmen thinly disguised) want. Their greatest fear must be a financially aware trustee who thinks to ask why a pension scheme should be buying a high cost, low return product.

As long as trustees are given this sort of advice, then those who invest in 'alternative' assets will always be viewed as creatures from a parallel universe where people say and think the opposite of what trustees are told to be the case. This is the scope of the problem which the various 'alternative' asset classes have to face when marketing to European pension funds, and particularly those in the UK. It is no coincidence that private equity and hedge funds, even when added together, account for just 1 per cent of UK pension plan assets.

Allocation levels

Nowhere is this rent in the time/space continuum seen more clearly than in the field of allocation levels. Swensen in his first book suggested that an allocation of less than 15 per cent to any asset class makes little sense. It cannot possibly make any difference to the overall return of the portfolio as a whole, and it is unlikely that an investor will be prepared to make a proper commitment of time and resource to do it properly. Subsequent discussion seems to have softened this slightly to about 10 per cent. Whatever the case, this means that there is not a single pension fund anywhere in the UK (or probably in Europe) that has a proper level of allocation to any of these areas. There was a tremendous fanfare of publicity, for example, when Hermes went to 5 per cent in each of private equity, hedge funds and property. What all of the press missed was the fact that this was still at least 100 per cent too little. If you are not prepared to put at least 10 per cent into each asset class, then you would do better to have no exposure at all.

Access and capacity are often given as reasons why such allocation levels are impracticable, but these are convenient rather than compelling. Let us take private equity as an example. It should be possible to commit at least $500m a year; indeed, there are may funds of funds and other investors who are looking to do at least twice this. For a pension fund investor, that would suggest an allocation of about $2bn, which means that one would have to be at least a $20bn fund before these became issues of any substance.

The future

Somebody once said that all you require to be happy in life is boundless optimism and a very short memory. Unfortunately, this is all too true in the case of the investment habits of European pension funds. Any idealistic belief that ultimately truth and reason must prevail is instantly assailed by the knowledge that this has not actually happened over the 30 years or so that the Yale Model has been in operation, and that nor do pension funds around Europe, even today, show the slightest interest in adopting a similar approach.

Logically one could argue that if given a choice between paying into an occupational scheme and not doing so then most individuals (certainly those under the age of forty) would actually be better off not doing so. This may strike some as an irresponsible thing to suggest, but the truth is probably that if the individual were to construct their own multi-asset class passive portfolio then it would probably perform better over the years than their pension fund, even after taking into account any possible benefits of tax exemption.

That would however be an acknowledgment of defeat and it is perhaps still too early for this. We must continue to hope that if the facts are placed repeatedly and cogently before trustees then eventually they will see the light. However, we must expect the fight to be long and difficult. LDI vendors might be seen rather as catholic priests selling indulgences for people's sins in the Middle Ages. They enjoy an entrenched position, and they are profiting from a belief system that they themselves have created and control. It is difficult to imagine a better business model.

The followers of 'alternative' assets might be seen rather as the original protestants who brought about the Reformation. At the moment we must content ourselves with creeping through the streets at night distributing subversive tracts, but sooner or later our movement will gain enough adherents that it will become adopted as mainstream, at least in some quarters. We have to believe this, no matter how improbable it may currently seem, because it is the only chance. Only then will the members of pension plans stand some significant chance of enjoying the sort of retirement they deserve. The alternative is despair.

References

References and Notes

1. Fraser-Sampson, G. (2007) 'Eppur si muove — Risk and alternative asset classes for pension funds', Pensions, Vol. 12, No. 2, pp. 82–87. | Article |
2. Swensen, D. (2000) 'Pioneering Portfolio Management', The Free Press (Simon & Schuster), New York.
3. Fraser-Sampson, G. (2006) 'Multi Asset Class Investment Strategy', John Wiley & Sons, London.
4. Source: Private Equity Intelligence.
5. Markowitz, H.M. (1952) 'Portfolio selection', Journal of Finance, Vol. 7, No. 1, pp. 77–91 and Markowitz, H.M. (1959) Portfolio Selection: Efficient Diversification of Investments, John Wiley & Sons, New York.