23.1 The French Money Market Fund Industry. How history has Shaped a Potentially Vulnerable Framework

Before starting our stories, let's first recall briefly how the MMF has become a major segment of the French asset management industry in the 1980 and 1990. This development owes much to three facts: (i) the legal restriction to the remuneration of sight deposits for nonfinancial corporations and households; (ii) the high level of real short-term rates, because of the “franc fort” policy in the aftermath of the infamous devaluations in the early 1980. These rates would remain high until the mid-1990 because of the Bundesbank's restrictive approach in the wake of German reunification of the early 1990, in spite of a quickly decreasing inflation; (iii) a very low taxation of income considered as long-term (LT) capital gain for holdings of 2 years or more.

MMF were first built as very short-term bond funds, the portfolio of which was purchased and sold back on an overnight basis, thanks to a put option, under a provision of the French Code Civil called “vente à réméré” (Art 1659), to fill the lack of repurchase agreement regulation. At the end of the 1980s, Certificate of deposits (CD's), commercial paper (“billets de trésorerie”), and T-bills (BTF) had become usual assets in MMF portfolios, between overnight repos and short maturities, govies, and corporate-floating rate bonds. As fierce competition reigned among banks¹ to retain investors, portfolio managers were eager to enhance their performance and especially to beat their mythic benchmark, the TMP (the money market overnight rate average: today's EONIA). And here started the paradox.

A large fraction of MMF was held by the treasurers of nonfinancial corporations, who used these funds as receipt for their cash positions, whatever their origin and destination. One part of these positions resulted from daily movements, predictable or not, and the other represented short- to medium-term hoardings, which back then were major contributors to firms' profitability, given the high level of interest rates. Institutional investors (insurance and reinsurance companies, pension funds) would also invest in these funds according to their in/outflows cycles. But from a strictly legal point of view, all these holdings had totally undefined maturities and were indeed redeemable with “today's value,” without any prior notice.²

Portfolio managers thus embarked in a nonclassical transformation business, in which their mightiest and smartest customers were asking for the higher returns, unavoidably involving assets with medium- or even long-term maturities, while retaining their instant exit opportunity. While this may look like a usual banking business, involving maturity transformation based on a statistical approach of depositors' outflows and sound liquidity ratio management, the major difference, however, is that the tradeoff between assets returns and balance-sheet soundness in commercial banks does not interfere with short-term (and especially sight) deposits collection.

MMF portfolio managers had, in fact, a limited number solutions to enhance their returns: (i) increase the average maturity of their assets, in order to get either higher fixed rates – provided that the yield curve will remain steep and that there won't be any rate hike by the Central Bank – or higher floating rates margins; (ii) downgrade the credit risk on short-term securities or medium- to long-term FRN for the reward of higher credit spreads; or (iii) build synthetic arbitrage positions through derivative strategies. Needless to say, except for the last one, any solution would surely weaken the sustainability of the fund in the face of massive outflows. Just for once, informational asymmetry would not be in favor of financial industry. Indeed, each party accepted a sort of moral hazard. But as far as markets cruised quietly with the comfortable cushion of an 8–10% risk-free rate, nobody worried.

23.2 The 1992–1995 Forex Crisis

In the year 1992–1993, Western Europe experienced several monetary crises, stressing Forex and Euro and domestic money markets as well. European currencies had been pegged within the EMS, the pre-Euro system in which the fluctuation of each currency versus any other member should not exceed ±2.25% from the central parity. Because of severe economic performance decoupling (mainly on inflation), these quasi-fixed exchange rates quickly became unsustainable. After the devaluation of the Lira and the scathing defeat of the British Pound by G. Soros (September 1992), the French Franc was in turn attacked by international speculation (such attacks would be repeated in February/march 1995 for more specific French reasons, but the boundaries of the system had been widened to 15%, so the effects on money market were less severe). Unlike their English counterparts, the French authorities decided to defend the sacred FRF/DEM parity by withdrawing liquidity from the money market and sharply raising the overnight rate of interest, in order to cut back the funding of speculators selling borrowed French Francs and eventually to punish them (see the peak in Figure 23.1). As a consequence, the yield curve suddenly reversed, as the crisis was seen to last only a few weeks, everyone betting on either a devaluation of the Franc or the defeat of speculators.

Now, what happened to MMF? Tracking a benchmark suddenly jumping from 10% to 13.5%³ was strictly impossible except for totally liquid funds (i.e., made only of overnight repos). For most funds, the “normal” rule dramatically reversed: the shorter the maturities of portfolio, the smaller their tracking error. Conversely, nonfinancial MMF holders, who were keen to browse MMF league tables day after day, didn't take long to exit from the laggards and invest in the best performing funds or, more often, directly purchase short-term securities. Such moves sparked the diabolical spiral of growing outflows striking the less liquid funds as their performance deviated from the benchmark.

As far as we know, no fund was shut down at this time, even though some were close to collapse. We guess the main reason was that banks who were at the same time promoting and managing the funds – independent asset management companies were still the exception – were eager to refund their MMF thanks to the cash they were collecting from the exiting companies by issuing Certificate of deposits.

This self-correcting mechanism notwithstanding, this crisis had shed light on the extreme vulnerability of the system in the face of a shock that was supposed to hurt only the benchmark yield and cause only performance disorder. Three years later, the author of the present chapter carried out an audit mission on the MMF recently launched by the Greek branch of a big international bank that comprised many medium-term debt securities among their assets. Unsurprisingly, the Greek colleagues listened politely to the above story, unbelieving it could ever strike Greece. What happened to their funds 10 years later, when the first Greek crisis burst, remains unknown. We guess, however, that like most fund managers in Europe, they had felt vindicated by the early success of the Euro, which was supposedly definitely dissipating the threat of monetary crisis within the Euro area – and had just kept purchasing medium-term private bonds, no matter the risk of liquidity.

With the introduction of the Euro, exchange rate fluctuations and therefore exchange risks within the Euro area have disappeared as expected, but as everyone is nowadays well aware, liquidity crisis stress may still spread across the banking and the MMF systems, and even faster. In fact, most self-corrective mitigations of the 1990s have receded, and several factors have aggravated the vulnerability of the credit market leading to the collapse of 2007–2008.

23.3 Four Mutations Paving the Way for Another Meltdown

The decade between the mid-1990s and 2006 saw one of major moves that would ultimately bring the greatest financial crisis since the 1930s. Among these, we retain as major determinants of the market stress four economic/financial evolutions:

• The promotion of home ownership, successively by Presidents Clinton and GW Bush, leading to a real-estate bubble in the US, thanks to the accommodative monetary policy of Greenspan/Bernanke and the easing of credit discipline, bottoming with the subprime boom;
• A regulation revival, in both the banking and the asset management industries, setting new requirements for banks capital (Cooke ratio) and severe assets and stockholders diversification ratios for mutual funds (notably in France);
• The progressive entry of the industrialized world into a low inflation/quasi-deflation era – due to globalization and new information technologies – implying ever lower interest rates, especially for “risk-free” assets. Unlike the Japanese nonfinancial corporations, institutional investors and savers across Western countries who had been used to comfortable nominal and real return on their fixed income portfolios were simply not ready to accept such declining yields and were thus be receptive to all kinds of strategies aimed to keep their passed returns alive;
• A bunch of financial innovation, especially in the field of debt, with derivative products and synthetic securities (CDS, collateralized debts obligation (CDO), collateralized loans obligation (CLO)) deeply modifying the origination/distribution/financing/refinancing production line for credit. Trying to assess the return distribution of such structured product implied also a real test of the fine line between risk and uncertainty.

Much has been written about the cascade of events that led from the burst of the American asset bubbles to the global credit crisis. Our contribution will thus only describe the case of dynamic MMF⁴ and how some of them ultimately were locked to prevent bankruptcy, as they couldn't meet massive withdrawals. What is not known by the public is how some dynamics MMF had progressively became complex, long-term corporate bond funds, in order to satisfy the demand for high return from addict investors, in spite of the huge decrease of short-terms rates, and how this transformation made them vulnerable to any freeze of the secondary market. Three special mentions must be made in this respect.

In the second half of the 1990, European banks were looking for equity or quasi-equity funding in order to match the requirements for capital (Cooke ratio, Bâle 1). As continental fixed income mutual funds, which were the richest, were not allowed to purchase equities, financial innovation was required to fill the gap. Having the same product construed differently (in terms of risk) by each party – and each party's supervisors – has always been a key mission of financial innovation, as a unique way to rebuild an otherwise broken confrontation of supply and demand. Infinite-duration subordinated debts (TSDI) were typically created under this rule. The trick to make them compatible with fixed-income funds regulation was the addition of a set-up provision, under which at a certain maturity (generally every other 5 years) the security would either be redeemed or see its coupon (or margin) increased by 75 or 100 bps. Although this construction was clearly a call option in the hand of the issuer, it was construed – and actively sold – as a put option in favor of the investor.⁵ The underlying assumption was, of course, that the funding market will go on without any disruption for at least the next decade, making it cheaper for the banks to issue new bonds rather than bear additional costs on the existing ones. Such securities were thus considered as ordinary medium-term bonds, maturing at the first step-up date, and offering however a larger spread than genuine fixed maturity securities. What should have caught investors' attention was the growing number of banks issuing such securities, at the same time creating a sort of critical rendez-vous for the credit market 5 or 10 years later. Put in another way, the premise of easy refinancing should work as a binary option: for all or for none. Thanks to this singular application of the extreme correlation principle, the so-called diversification benefit was nil.

A second mention must be given to the world of securitization (asset-backed securities, CDOs) and the efforts made by financial scientists to model the credit risk for such instruments. As banks were less and less willing to retain the loans in their own balance sheet to match prudential constraints, securitization became the ultimate step of the disintermediation of financing the economy. But as the credit risk burden was only displaced, two unknowns remained when assessing the probability of default of such collateralized instruments. (i) Whereas one can correctly estimate the distribution law of one single default, anticipating the modification of correlations inside a cohort of similar primary loans in a spreading crisis remains highly uncertain, as mentioned above. (ii) For multi-tranch CDOs like those including subprime, it is therefore impossible to assess how the lack of guaranty will roll down toward the tranches with higher rating, once the defaults have started to burgeon.

Neither the financial institution packaging such securities nor the rating agencies paid sufficient attention to those underestimated risks, as the key hypothesis was that there would never be a real-estate crash in the US, even though recent Japanese history was a perfect counterexample.

The third point deals with the regulation provisions that were supposed to warrant the sustainability of the financial system. In another version of the old debate about the reflexive effects of mandatory stop losses, diversification ratios may introduce a destabilizing effect in the market by forcing the portfolio manager and/or the investor to sell in an adverse market. Under French regulation, no investor may hold more than 10% of a fund. This rule creates real hurdles when launching a fund, but that is another story. Now, just consider that a rumor hits a certain issuer (bond, perpetual subordinated, CDO), the security of which is present in your portfolio. One of your fund holders, too cautious and/or too worried, decides to sell his shares, representing 9.5% of the fund. The next investor, with an initial holding of, say 9.2%, has now a 10.2% shares, exceeding in turn the limit and therefore forcing him to sell. Whereas he may sell only the exceeding fraction, he may also be tempted to sell more, and so engage the downsizing spiral for the fund. On the other hand, as the capital of the fund decreases, the manager has to sell liquid securities in order to pay the outflows. But as the total portfolio decreases, he must also, at some point of the process, sell private securities, which now represent more than 5% of the portfolio, which is the upper limit for private assets. This is the common way that causal chains of illiquidity start when fund managers suddenly facing self-reinforcing constraints are deprived of their ability to rationally decide.

23.4 The Subprime Crisis Spillover. How Some MMFs were Forced to Lock and Some Others Not

Now let us consider that this was not one single rumor, but real business crashes (Bear Stearns June 2007; IKB July 2007; Northern Rock Sept 2007) with many projections for similar companies, as the subprime meltdown suddenly appeared as a global affair. In fact, the credit market (corporate bonds, bank-subordinated and ABS) had been virtually illiquid since the beginning of January 2007. The author of this chapter had taken over, in September 2006, the management of a small dynamic fund (less 100 M Euros), the portfolio of which comprised more than 50% of subordinated and traditional corporate bonds. The decision was quickly taken to sell the totality as the only safeguard for this small independent asset management company against the lethal script described above. But, as the credit market collapse loomed, it actually took more than 6 months to sell the half-dozen subordinated banks securities and another eight BBB corporate bonds.⁶

On August 2, 2007, Oddo, France's largest independent asset manager, and BNP Paribas, one of Europe's largest bank, decided, with the agreement of French regulator AMF, to lock several dynamic MMFs because of severe NAV losses due to illiquid ABS linked to the US real-estate market.⁷

Locking an MMF is like closing a bank position in front of a bank run. Depositors' panic may spread and have dramatic consequences, as was seen in Iceland. In the 2007–2008 crisis, different scenarios emerged: pure bankruptcy (Lehman), total nationalization (Northern Rock; Dexia; AIG), public bail out (French banks), temporary lock-up of funds, etc. Although it was strongly discussed, the “too-big-to-fail” principle fully applied (except for Lehman, and most probably because of the apocalyptic reaction to Lehman's failure). But for smaller struggling companies (banks and asset management), the reputational effect of repurchase suspension would put a tragic end to their effort to survive, let alone grow.

At the other end of the risk–reward range of AM products, the multi-managers' alternative asset management segment (MMAAM: funds of hedge funds) took a death blow in 2009 from the “gates” it had to impose on subscribers. In their own style, MMAAM was another complex system, involving, between the final investor and the assets and derivative markets, two levels of interdependent fund managers.

23.5 Conclusion. What Lessons can be Drawn from these Two Tales?

As a conclusion, we are not supposed to dispute in this pages the relevance of the official responses to the extreme events encountered during the 2007–2008 banking crisis, although there is some evidence that public interventions to avoid bankruptcies and the clemency of regulators when prosecuting professional misbehaviors have encouraged the astonishing revival of foolishness in financial industry after 2010. However, three essential lessons may be drawn from these short and sharp cold–hot cycles.

The first lesson is in fact a theoretical question: considering the last 30 years of market history, one can ask: what is the exact meaning of extreme? In other words, does “extreme” refer to the frequency of severe events or to the degree of severity of such events? From Bachelier, at the dawn of the twentieth century, to Mandelbrot, mathematicians have worked hard to assess the distribution laws of periodic returns and the size of their tails (thin or thick). But haven't we entered a new era, where extremely severe episodes burst with a sort of regularity, from the October 1987 daily crash to the 2010–2011 European debt crisis, the violence of which put the survival of short-sighted companies and mutual funds brutally at stake? A positive answer would introduce a major shift from stochastic to more deterministic binary approach of risk across all asset classes, with a new recognition of time and cycles.

This observation directly leads to our second operational lesson. Extreme situations, we think, are those in which certain categories of operators have no choice but to do things that rationally should not be done, thus contributing to strengthen the global irrationality of the situation. That's why, even though extreme risks may not be foreseeable, managers should always try to mitigate them by ensuring that a portfolio has kept a sufficient reserve of flexibility, albeit at the inescapable expense of return. What is obviously true for the diversification principle, which requires to remain totally efficient when correlations rises asset classes with low (or even no) risk premia but long time statistical independence, is all the more true for the liquidity risk, which can be covered, to a certain extent, thanks to securities that are permanently bought by the Central Bank and through a very careful management of short maturity asset roll-over.

The third lesson is a trivial one, although widely unlearnt: “things are not different this time” however sound the market looks. The American philosopher, George Santayana, once wrote: “Those who cannot remember the past are condemned to repeat it,” a quote often transcribed as “Those who do not learn history are doomed to repeat it.”

Further Reading

1. Galbraith, J.K. A short story of financial euphoria. USA: Viking Penguin; 1990.
2. Kindleberger, C. Mania Panics and Crashes. New York: Basic Books; 1978.
3. Longin, F., Solnik, B. Extreme correlation of international equity markets. The Journal of Finance 2001;LVI(2). p. 649–676
4. Minsky, H.P. The Modeling of Financial Instability: An introduction. Modeling and Simulation. Proceedings of the Fifth Annual Pittsburgh Conference 5; 1974.
5. Minsky, H.P. The Financial Instability Hypothesis. Working Paper No. 74 ; May 1992.
6. Ninet, J. Les déterminants récurrents des bulles financières. Economie Appliquée. T 57 2 ; 2004.