It is widely, if not universally, agreed that the world is experiencing a period in which intentions to save tend to outrun intentions to invest. Alan Greenspan and Ben Bernanke, who is Chairman of the US President's Council of Economic Advisors, are among those who have pointed to this probability. If economic policy did not counteract this, the result would be a world recession. Fortunately this has so far been avoided or at least postponed by the easy fiscal and monetary policies.

The result is that all major economies have large budget deficits and interest rates are exceptionally low in both real and nominal terms. The world is awash with money and this has driven up the price of assets, be they bonds, houses, other properties or equities. This has had the desirable result of depressing the wish to save, but it has created a grave problem for pension funds. It is likely that returns from investing in these asset classes, which have been the traditional universe for pension funds, are likely to be extremely poor.

This is an extremely uncomfortable prospect for the sponsors of pension funds and for their advisors and fund managers. This discomfort creates problems. Telling plan sponsors that they should increase their pension contributions and hold cash in the fund is a rational and intellectually honest response, but it puts businesses and careers at risk. The demand for other answers is therefore massive, and demand produces its supply. In financial services, even more than in most industries, it might be said that "where there's a will there's a way." There is of course the risk that the old adage of Addison Mizner will be even more apposite. As he wrote in his cynic's calendar "where there's a will there's a lawsuit".

The two most popular routes for escape from the probability of low medium-term returns from traditional assets are to invest in private equity or hedge funds. In this article I shall concentrate on the latter and I must declare an interest. I number many hedge funds among my clients and I am therefore likely to see virtues in them. My fear is that if any of them read this article, they will not think that I have found enough.

Hedge funds undertake a much more varied range of activities than traditional long-only managers. It has been said, unkindly but by members of their own community, that "hedge funds are not an asset class, they are a charging mechanism". Hedge fund managers charge a lot. It is true that their charges are generally linked to performance. But if all funds were "hedge" the total payments made by investors would rise. In aggregate, therefore, they can only benefit investors if they either outperform the "unhedged" or improve aggregate returns.

There is a case that they can do both. (I should, however, remind you here of my bias.) But there must be grave doubts as to whether they can do so. (Am I biased enough?) I see the major threats to the positive case as being first, whether the "industry" becomes large and second, whether charges are not so large than any aggregate benefits fall on the managers rather than on their investors.

One type of hedge fund is "long/short equity". These differ from long-only funds in being able to seek absolute or relative outperformance by selling over-priced shares short, rather than simply not holding them. The case for this type of fund is that if some managers are capable of superior share selection, they will be able to create more value from this if they can short stocks. Another way of putting this would be to point out that information costs money and that if fund managers can go short as well as long, the information that they have will not cost any more, but will have greater value.

As I see it, there are three problems with this. The first is whether superior fund managers can be selected. The second is whether their superiority will survive the growth of the industry. The third is whether the cost to the investor will be 100% or more of any added value.

In an efficient market, a group of investors who managed funds which were in aggregate market neutral would achieve in aggregate the same return as would be available on cash, before expenses. Such funds would then be an expensive way of holding cash.

I think there are three conditions necessary to avoid this. Firstly, the market must not be fully efficient by enough for past performance to be a guide to the future. Secondly, the money that these managers manage must not grow to the point where the market becomes efficient. Thirdly, the managers must leave something for the investors.

But long/short equity is by no means the only strategy open to hedge funds. Another is investing in types of assets which have traditionally been the main activity of banks.

It is obviously impossible to make a return above the risk-free rate without taking risk and only necessary to say so because the point seems so regularly overlooked by optimists. The way that banks remain in business is by taking two sorts of risk, which have the advantage that they are probably less risky taken together than they are in isolation.

Banks take credit risks by lending money to people and companies who can go bankrupt. If they select them well, the banks will make more money by charging a premium over their cost of borrowing, than they will lose from the periodic bankruptcies. Banks also buy bonds. This is a different sort of risk. They usually make more money from holding bonds, even "risk free" ones, than they have to pay out to their depositors. But as the interest on these bonds is fixed until they mature, they will lose if the cost of money on deposit rises.

These two risks are nicely balanced. The economy goes up and down; in good times interest rates go up and bankruptcies go down, and vice versa. Thus the banks tend to make more money from holding bonds in bad times and less money from lending.

In recent times, however, banks have increasingly sold off their credit risks to other investors, with hedge funds being very important. This process has sparked off a whole new product for the securities industry called CDOs or CLOs. These are Consolidated Debt or Loan Obligations. These are just packages of debts and they can be sold off in tranches, so that the buyer can take the top risk, which occurs when any of the borrowers stop paying, or the less risky bits, which will only have problems if lots of borrowers are in trouble.

One feature of this process which has, I think, received too little attention is that when profits are made by banks by taking credit risks, these profits are liable to corporation tax. But if the profits are made by pension funds, either by direct investment or via hedge funds, no such tax is payable.

If, as seems likely, part of the business of hedge funds consists of reducing governments' revenue from corporation tax, then hedge funds will be able in aggregate to improve aggregate returns. To do this they need a patsy and that patsy will be the tax system.

It is generally believed that pension funds were robbed by Gordon Brown when he abolished ACT. In a small way at least, hedge funds might provide a route for pension funds to get their own back.

Banks, as a result of regulation, are implicitly and in some countries explicitly guaranteed by governments. This lowers their borrowing costs, which in turn makes it unlikely that they will not continue to be the main intermediary between lenders and low risk borrowers. But the tax advantages will probably be the dominant long-term determinant of who undertakes the more risky elements of lending, because these require a high element of equity, where the cost is heavily influenced by tax.

We should therefore expect the risky elements of lending to be increasingly owned by investors who, unlike the usual practice of banks, "mark to market". This is likely to make risk margins more volatile.