Global Investment Outlook 2012 — The Sentence That Repriced a Continent

The euro-area crisis ended in July 2012, and the mechanism by which it ended is one of the most instructive events in modern financial history.

A central bank announced it would purchase the sovereign bonds of stressed member states without limit, subject to those states accepting a formal programme. Peripheral yields fell immediately and continued falling for years. The programme was never used.

The instinctive reading — that markets are irrational and respond to talk — is wrong, and the correct reading is far more useful.

What actually happened is a shift between equilibria. The Global Investment Outlook 2011 sets out the self-fulfilling run: doubt about rollover raises yields, higher yields make the debt genuinely harder to service, which justifies the doubt. That loop has two stable outcomes for the same underlying fundamentals:

• A good equilibrium where investors expect rollover, yields stay low, debt stays serviceable, and the expectation is validated.
• A bad equilibrium where investors doubt rollover, yields rise, debt becomes unserviceable, and the doubt is validated.

Both are internally consistent. Nothing about the economy determines which one obtains — only expectations do.

A credible unlimited backstop eliminates the bad equilibrium, because it makes step three impossible: yields cannot rise to unserviceable levels if a buyer of last resort stands behind them. Once the bad equilibrium is unavailable, the good one is the only place to be, and the market moves there without the backstop being touched.

This is why the promise did not need to be kept in order to work. It needed only to be believed.

The year's second theme is what a fourth year of near-zero rates did to portfolio construction. What began as a cyclical response had by 2012 lasted long enough that institutions with fixed return requirements stopped treating it as temporary and rebuilt around it — a structural change whose consequences the archive tracks through the following decade.

Multiple equilibria, and why they matter

The idea is unfamiliar outside economics and is worth developing, because it applies far beyond sovereign debt.

In an ordinary market, fundamentals determine price. Better earnings, higher value. There is one answer, and analysis is the search for it.

In a market with self-reinforcing expectations, the outcome depends partly on what participants expect the outcome to be. That can produce two or more self-consistent answers for identical fundamentals.

The classic case is a bank run, which the Global Investment Outlook 2008 describes at length. A solvent bank funded by demandable deposits has two equilibria: everyone stays and it is fine, or everyone runs and it fails. Both are rational individually. Deposit insurance eliminates the bad one — and, crucially, deposit insurance is rarely paid out, because its existence prevents the run it insures against.

The 2012 sovereign case is structurally identical:

Bank run	Sovereign run
Depositors may withdraw	Investors may refuse to roll over
Withdrawal forces asset sales at loss	Refusal forces yields up
Losses justify the withdrawal	Higher yields justify the refusal
Deposit insurance removes the bad equilibrium	Unlimited backstop removes the bad equilibrium
Insurance rarely paid	Programme never used

Three consequences follow, and all of them are practical:

Fundamentals do not fully determine price in such markets, so analysis that stops at fundamentals will misprice them in both directions.

Policy can be extraordinarily powerful at very low cost when it changes which equilibrium obtains rather than trying to change the fundamentals.

And the reverse holds. A backstop that stops being believed restores the bad equilibrium instantly, with no change in fundamentals. The value of the commitment is entirely its credibility, which is a stock that can be depleted — the concern the Global Investment Outlook 2022 examines when the same institutions faced an inflation shock.

The announcement worked because it made the feared outcome impossible, not because it made investors feel better. Sentiment follows structure, not the other way around.

Why conditionality was the whole point

The programme was conditional: a member state had to request support and accept a formal adjustment programme before purchases could occur. This was widely read at the time as a political constraint weakening the commitment. It was closer to the opposite.

Consider an unconditional promise to buy any member's bonds without limit. Its problems are immediate:

• It creates moral hazard. A government whose borrowing costs are guaranteed regardless of behaviour has weakened incentives to control them.
• It is monetary financing of governments, which the union's founding framework prohibits — so it would face legal challenge, which is itself a source of doubt.
• It is therefore not credible, because market participants would correctly doubt it could survive contact with law and politics.
• And an incredible backstop does not eliminate the bad equilibrium, which means it does not work at all.

Conditionality resolves each problem. It preserves the incentive to adjust, it distinguishes the action from unconditional financing, it gives the commitment a defensible legal basis — and by making the promise credible, it makes it effective.

This generalises into a principle worth carrying:

A promise strong enough to solve the problem but too strong to be believed solves nothing. Constraints that make a commitment survivable are what make it work.

The Global Investment Outlook 2013 shows the same principle from the failure side — a communication that was not understood as intended produced a violent repricing, because what matters is not what is promised but what is credibly received.

The reach for yield turns structural

By 2012 policy rates in the major developed economies had been near zero for four years. The duration is the important fact, not the level.

Why time changes the nature of the adjustment:

• In year one, an institution with fixed obligations waits. Rates are low, this is temporary, the return assumption holds.
• By year four, waiting has become a decision with a compounding cost. The institution must either change its obligations, change its return assumption, or change its portfolio. The first two are usually impossible for legal or contractual reasons.

So portfolios changed, in consistent and predictable directions:

• Down the credit spectrum, into lower-rated corporate and emerging market debt.
• Out the duration curve, accepting more interest rate sensitivity for more yield.
• Into illiquid assets — private credit, real estate, infrastructure — where the illiquidity premium substituted for the vanished risk-free yield.
• Into complexity, where structure rather than underlying risk generated the additional return.

Each of these is a rational response to the constraint. The problem is that they were rational for everyone simultaneously, which means the same assets were bid by the same institutions for the same reason.

That is a correlation the individual portfolios do not display. A pension fund's private credit allocation looks uncorrelated with its high yield allocation. Both are the same trade — a search for return in a zero-rate world — and both reprice together when rates change, which is exactly what the Global Investment Outlook 2022 documents a decade later.

The archive's Private Markets Outlook 2016 tracks the illiquid leg of this reallocation specifically, and the Fixed Income Outlook 2019 tracks the credit leg.

Reading central banks becomes a discipline

2012 marks the point where central bank language became a first-order driver of asset prices rather than a commentary on them.

Why the shift happened:

• With policy rates at zero, the rate itself carried no information. It could not go lower, so it could not signal.
• The remaining instruments were expectations-based — asset purchases, and statements about the future path of policy.
• Both work only through what markets believe, which means the communication is the policy, not a description of it.

This created a genuine analytical problem. Interpreting a central bank statement requires distinguishing:

• A change in the reaction function — the rule linking conditions to policy. This is major and reprices everything.
• A change in the forecast — same rule, different expected conditions. This is moderate.
• A change in emphasis or wording with neither of the above. This should reprice nothing and frequently repriced a great deal.

Most market reactions to central bank communication are reactions to the third category, which is a persistent, structural source of mispricing rather than an occasional one.

The practical test worth applying to any policy statement is to ask what conditions would now produce a different action than they would have produced a week ago. If the answer is none, the reaction function has not changed, whatever the tone suggests. The Global Investment Outlook 2013 is the case study in what happens when that question is answered wrongly by the whole market at once.

The half-built union

2012 also began construction of the architecture the Global Investment Outlook 2010 identified as missing — and it is worth being precise about which parts were built, because the gaps persisted for years.

A banking union has three legs, and they do different jobs:

Common supervision — one authority overseeing the major banks, rather than national regulators with national interests. Agreed in 2012 and implemented within two years. This was the easiest leg because it transfers oversight rather than money.

Common resolution — a single mechanism and fund to wind up failing banks, so a bank failure does not fall on its home sovereign. Agreed later and built with a fund that took years to accumulate, meaning the leg existed in law before it existed in capacity.

Common deposit insurance — a guarantee that a euro in a deposit is worth a euro regardless of which member state the bank sits in. Not built. It remained under discussion for more than a decade.

The third omission is the one that matters most for the mechanism this report describes, and the reasoning is direct. Redenomination risk, per the Global Investment Outlook 2011, transmits primarily through deposits. A depositor who fears their euros might become something else moves them to a bank in a member state where that fear does not apply. Common deposit insurance is precisely the instrument that removes that incentive, because it makes the location of the bank irrelevant to the safety of the deposit.

So the union built the leg that prevents bank failures and skipped the leg that prevents bank runs — which is a reasonable political sequencing and leaves the original fragility partially intact.

Why this stopped being priced: the 2012 backstop worked so completely that the underlying structure was no longer tested. A risk that is not tested generates no price, which is not the same as a risk that is absent — the archive's most persistent finding, and the reason the Europe Investment Report 2020 treats the eventual fiscal response as the delayed completion of work begun here.

What an allocator could act on

Ask whether a market has one equilibrium or several. Where outcomes depend on expectations — bank funding, sovereign rollover, currency pegs, run-prone structures — fundamentals do not determine price, and analysis that stops there will be wrong in both directions.

Value a backstop by its credibility, not its size. An unlimited commitment nobody believes is worth nothing; a bounded one that is fully believed can eliminate a bad equilibrium entirely. Ask what would have to happen for the promise to be broken.

Read constraints on a commitment as sources of strength. Conditionality made the 2012 programme legally survivable and therefore believable. A promise engineered to be unbreakable is usually a promise nobody has tested.

Aggregate reach-for-yield exposures as one position. Lower-rated credit, extended duration, illiquid assets and structural complexity were bought by the same institutions for the same reason, so they carry a correlation that shows up in no risk report and appears only when rates move.

Test policy statements against the reaction function. If no set of conditions now produces a different action than before, nothing has actually changed. Most large reactions to central bank language fail this test.

Note that the absence of a crisis is not the absence of risk. The 2012 programme worked so completely that the underlying fragility — a currency union still lacking a banking union in full — stopped being priced. A risk removed by a credible promise remains present, held in the form of that promise.

What 2012 established

• A credible backstop eliminates a bad equilibrium without being used, which is why the programme worked while never purchasing a bond.
• Multiple equilibria are real in expectations-driven markets, and policy can move between them at almost no cost.
• Conditionality made the commitment credible and was therefore the enabling condition rather than a weakening one.
• The reach for yield became structural at four years of zero rates, creating a hidden common factor across apparently unrelated allocations.
• Central bank communication became a primary price driver, making the distinction between a changed reaction function and a changed tone a practical necessity.