With all the fuss about hedge funds and private equity these days, I think it's worth cataloging in simple terms the multiple levels of "agency costs" associated with leveraged investment funds. [1]

1. Agency costs imposed by fund managers on investors
2. Agency costs imposed by fund investors on banks and other creditors
3. Agency costs imposed by bank managers on bank shareholders, depositors and governments
4. Agency costs imposed by investment managers on institutional investors
5. Agency costs imposed by institutional investors on governments and the public at large

1. Agency costs imposed by fund managers on investors

   As is widely discussed, many hedge-fund managers have a "two and twenty" fee structure, meaning they take as fees 2% of the funds they manage under any circumstances, and 20% of any profits achieved. Because fund managers see the upside of gains but not the downside of losses, a rational self-interested fund manager will take more risks than her investors would if they were managing their own money. Also, fund managers often compete on the basis of short-term apparent performance. A rational, self-interested fund manager may prefer a "get rich quick" strategy to a long career. Such a manager might take positions that trade high immediate returns for large future risks. Credit default swaps are an ideal vehicle for this sort of thing. A "get rich quick" fund manager might also aggressively value illiquid investments on the fund's books, creating high phantom returns that cannot be realized when the positions are liquidated. Conversely, she might take-on "off balance sheet" liabilities, inflating the apparent value of the fund. Carefully written "derivatives" permit parties to assume contingent liabilities in ways that hide the leverage of the fund. [2]

2. Agency costs imposed by fund investors on banks and other creditors

   Limited-liability creates a potential conflict of interest between investment funds and their creditors. If a fund is heavily leveraged, fund investors can reap large rewards by assuming risky positions with the understanding that if those positions go sour, a large fraction of the cost can be shifted (via actual or threatened bankruptcy) to the fund's creditors. I've described this at length in a previous post.

3. Agency costs imposed by bank managers on bank shareholders, depositors and governments

   Keynes famously wrote in 1931 that "a 'sound' banker, alas, is not one who forsees danger and avoids it, but one who, when he is ruined, is ruined in a conventional and orthodox way along with his fellows, so that no one can really blame him." [3] There is no reason to believe that bankers are any less sound today then they were in the 1920s, or the 1980s. Banks earn money in a highly competitive environment by lending money, and bank managers are evaluated by the profitability of their operations in relation to those of peers. Leveraged investment funds have become a breathtakingly large clientele. They hold trillions of dollars of equity which they are in the habit of borrowing against aggressively. Since the mid-nineties, the scale of hedge and private equity fund investment has grown astronomically, and few banks have been burned by extending them credit. Suppose that after the late 1990s when the Long Term Capital Management famously blew up, a bank manager decided that future lending to investment funds would be done conservatively and only on the basis of extensive due diligence. How would her performance have compared to that of his peers who lent more freely? A rational, self-interested bank manager may well conclude that her best strategy with respect to the huge, lucrative investment fund market is to "see no evil" with respect to systemic risks. Rational managers would diversify their exposure among many funds, to reduce fund-specific risks, but would want to aggressively pursue business volume in the sector as a whole. This is a classic "tragedy of the commons". By competing well for volume, bank managers exceed performance benchmarks and earn bonuses. But they also diminish the cost and increase the quantity of leverage available to investment funds, magnifying systemic risks. Should a "meltdown" occur, individual bank managers will point to their industry-standard, state-of-the-art risk management practices and demand safe harbor. A few managers with fraudulent books or eggregiously risky positions will be tarred and feathered. The rest will keep their bonuses, while bank shareholders, bank depositors, and eventually taxpayers eat their losses. [4]

4. Agency costs imposed by investment managers on institutional investors

   Like any other sort of investment manager, the interests of those who manage funds for pensions, university endowments, and charitable foundations may diverge from the interests of their diverse clientele. In particular, rational, self-interested managers may determine that pursuing peer-competitive short-term gains is wiser than carefully managing the long-term risks of fund stakeholders. Managers of pension funds, for example, may "chase high returns" via risky, leveraged investment funds, hoping to maintain or achieve "fully funded" status so the plan sponsor can avoid transfers to, or pull cash out of, the fund. In this case, there are two levels of agency cost: the managers who try to please their employers at the expense of fund beneficiaries by implementing the strategy, and the plan sponsors themselves, who impose risks on beneficiaries to extract cash or avoid future payments.

5. Agency costs imposed by institutional investors on governments and the public at large

   It is not only self-interested managers or pension-raiding corporations who impose agency costs on others. Institutional investors that increasingly invest in highly leveraged investment funds don't do so only so fund managers can look good. Institutional investors impose agency costs on the general public, when institutions important to the public take risks whose benefit accrue to direct stakeholders but whose costs would be shared by the community at large. Suppose a university chooses to invest its endowment in highly leveraged investment funds in hopes of receiving outsized returns, which it will use to fund expanded programs, higher salaries, and other good things. The improved endowment returns benefit the university community much more than it does the general public. A catastrophic loss of endowment funds, however, imposes very large costs to the surrounding community, which may have to underwrite a bail-out of some form if important programs are threatened. A rational, self-interested university might choose an investment portfolio which provides high, steady returns in exchange for assuming a small but significant risk of catastrophic loss. In this way, the university community enjoys enhanced programs and salaries under the most probable scenarios, and forces public intervention on its behalf under less favorable scenarios. This sort of strategy can be rationally adopted by universities, important charities, pensions, any sort of institution willing to gamble that, within its own community, it is "too big to fail".

[1] "Agency costs" is economist-speak for the damage done by conflicts of interest between decision-makers and those on behalf decisions are made. If the CEO of a firm acts to provoke an unsustainable spike in his company's stock-price to enhance his annual bonus, the costs borne by stockholders (the inflated bonus, damage to the firm if the CEO's actions fail to maximize is long term value) are agency costs. Other examples of agency costs are political corruption and doctors who over-recommend lucrative procedures to their patients.

[2] Again, credit default swaps are a good example. When it can be argued that a CDS is a simple "loan guarantee", the writer of "credit insurance" may not be required to account for its position as a liability, creating the appearance that the "premiums" are pure profit. I have no idea how widespread this is in practice, but it is a very large loophole in theory.

[3] Keynes quote from "Consequences to the Banks of a Collapse in Money Values", 1931. Hat tip to Calculated Risk, writing on Angry Bear.

[4] Note that bank managers can take profitable risks even when they appear to be shedding risk. When a bank purchases "credit insurance" in the CDS market, apparently the bank is reducing risk, and accepting a fixed cost to do so. But just as writers of credit insurance may seek to keep their positions off-balance-sheet, purchasers would want their positions on-balance-sheet, offseting the assumed of the insurance with an asset. Given a high willingness by investment funds to sell credit insurance, banks may be able to purchase this insurance at unduly cheap rates. When credit conditions change and banks revalue their assets, they can book a profit by upping the book value of their CDS positions from cost to fair-value. In a systemic crisis these positions may turn out to be worthless, as the highly leveraged funds that sold the insurance go bankrupt. (It's oversimplifying, but not inaccurate, to suggest that the banking system is selling insurance to itself by offloading risk to leveraged investment funds. The banks expect to be made whole by the very same clients whose inability to pay off loans may trigger their need to be made whole. Individual banks are lending to and buying insurance from different parties. But banks as a whole are doing much of their risky lending to the investment fund sector, and buying much of their insurance from that sector.)

This is an elaboration of a comment I posted in response to an Economonitor post.