By Taimur Ahmad

Seven years ago, former US Treasury Secretary Lawrence Summers proposed what, at the time, seemed to many people a radical idea.

Summers, then president of Harvard University, suggested in a lecture that a portion of the wealth being rapidly accumulated worldwide by central banks – chiefly in the emerging markets through their foreign exchange holdings – should be invested in more productive and higher-yielding assets, rather than ploughed into low-yielding US Treasury bills.

“6% would not be an ambitious estimate of what could be earned” by investing reserves domestically in infrastructure or in a “fully diversified way” in global capital markets, he told an audience at the Reserve Bank of India.

Accumulated reserves were far larger than what could be seen reasonably as self-insurance against adverse shocks, Summers said. Moreover, the risk composition of assets in which these “excess reserves” were invested likely implied a “zero real return measured in domestic terms”. Simply put: a “substantial cost”.

The total central bank reserve pool he was referring to in early 2006 was a staggering $4.3 trillion.

Today, it’s $11 trillion. It was spawned and fed in recent years by vast global external imbalances between surplus and deficit countries and by sustained high commodity prices. In the last year alone it grew by $725 billion, according to IMF data.

What Summers didn’t know when he gave his address in Mumbai was that a financial crisis would strike a year later which would snowball to engulf the entire global financial system and which would ultimately leave developed economies ravaged by an explosion in public sector debt and facing sovereign crises of their own.

But what was perhaps even less foreseeable was that official sector demand for US Treasuries would nevertheless continue unabated – even after its sovereign debt lost its triple-A rating, after the Federal Reserve resorted to three rounds of emergency bond buying and despite the negative real yields on those securities.

Doomsayers had long predicted an investor revolt against the US Treasury market. Yet government bonds were boosted by a global flight to safety; and investors, in particular foreign central banks, have not shied from piling into US dollar-denominated assets.

“The markets and the activities of reserve managers have been very uneven, in large part because the environment has been very disorienting,” says Terrence Keeley, head of the official institutions group at BlackRock, the world’s largest asset manager. “Given their preoccupation with credit and default risk, they have even bought negative yielding securities.”

Times are changing

Today, the debate is heating up anew over how the central banks of major surplus nations, including in Latin America, should manage their reserves – as uncertainty looms over the global economy as well as over the returns offered by conventional investment strategies.

BlackRock estimates that historically low nominal rates in the US, UK and Japan are costing the holders of $11 trillion in foreign exchange reserves some $250 billion a year versus pre-crisis levels.

Many central banks in emerging markets – some of which have relied on foreign exchange reserves to generate up to 80% of their total income – are now facing “unsustainable financial trajectories,” the asset manager says.

Just like pension plans, insurance companies and other institutional savers, official sector funds are being forced to change their investment strategies to replicate the historic returns they once earned just from government bonds.

In an interview with LatinFinance in March, Yi Gang, head of China’s $3.44 trillion State Administration of Foreign Exchange (SAFE), the world’s largest reserves holder, expressed his concerns over the deteriorating financial performance of Chinese state wealth.

Yi, deputy governor of the People’s Bank of China, resolved to press ahead with efforts to diversify his country’s foreign currency holdings into higher yielding assets, including emerging market currencies and bonds. He said expectations of continued poor returns from developed markets had prompted the central bank to “continue to diversify our reserve investment”.

“Everybody knows that recently [quantitative easing] policy has driven down the interest rate so that bond yields have decreased for years. That means a lot of challenges if you invest in this market,” he said. “We really have to have a diversified and balanced portfolio.”

The People’s Bank of China, like many other central banks, is increasingly investing in a wider range of non-core and riskier assets, including Latin and other emerging market sovereign debt and equities, as yields on developed world sovereign bonds remain subdued.

The growing presence of official sector capital in private markets represents a shift in the global economy – and one with significant implications for asset valuation, financial stability and growth.

That shift, if it happens meaningfully, could also represent one of the most profound opportunities yet for the capital markets of Latin America and other emerging regions – in particular for the development of local bond markets, which depend critically on central bank foreign exchange reserve allocations.

Capital markets promise

“The big capital markets innovation in the waiting, that would be the greatest source of capital markets reform and change, is central bank reserve allocation to the emerging markets,” says Ousmène Mandeng, a former IMF official who was, until recently, head of public sector advisory at UBS.

Central banks could help by providing “anchor money to allow more liquidity and help the development of local capital markets,” he says. “It’s a sensible argument to make: a lot of currencies are so attractive and liquid because central banks are investing in them. That’s certainly true of the main markets and that logic could be applied to the emerging markets as well.”

This would require central banks moving beyond the status quo in the international monetary system and towards a multiple reserve currency system– incorporating leading emerging market currencies including those of China, Brazil, and Mexico, among others. Chile’s central bank and the Reserve Bank of Australia, for example, are among a handful of China’s trading partners to have already diversified a portion of their reserves into renminbi.

Mandeng advocates establishing a framework, similar to central banks’ existing gold sale agreements, by which monetary authorities constrain allocations to new reserve currencies. The limit could be 15 percentage points of total foreign exchange holdings over the next five years, for example.

“This would help mitigate undue volatility of the main reservable and new reservable asset markets, help guide market expectation, support desired orderly reserves diversification and provide critical support for the development of local bond markets,” he says. “Central banks would be guided, similar to the central bank gold sales agreements, by making purchases of new currencies not in excess of an agreed limit and during an agreed time interval.”

Such a mechanism could help develop long-duration, local currency fixed income markets in Latin America, which – despite important progress over the past decade – are still small and illiquid by international standards.

The potential for such development is not lost on Yi. Emerging markets as an asset class are not large or deep enough for meaningful investment by official investors. But, he says, strides already taken in developing Latin America’s bond markets are “very positive”.

“I think [the] emerging markets asset class in the future will deepen its scale and volume and liquidity will be better. It has tremendous potential though right now it’s not very large.”

How much is too much?

Today, there’s a growing belief among central bankers themselves that reserve levels have surpassed useful limits and that the cost of maintaining them is now too great. This raises the possibility of a greater allocation of reserves as investable assets.

Summers’ argument hinged on the notion of excess reserves, which he put at $1.5 trillion in 2006. This raises the question: what is the appropriate level of foreign exchange reserves for a nation to hold? One theory, the Greenspan-Guidotti rule for the adequacy of reserves, says they should equal external debt maturing within 12 months; that is, enough to cover a sudden reversal in short-term capital.

For China, whose reserves stockpile is $3.44 trillion, the answer is clear. Says Yi: “We already have large enough reserves.”

Safety, liquidity and return, he says, are the three guiding principles of reserve managers – in that order. Reserves should first be about self-insurance against shocks. “The most important function of foreign exchange reserves is increasing confidence of the market,” says Yi. “If you pursue safety and liquidity and a little bit of return, the US Treasury market is a major market.”

But, he adds: “For the reserves it is not the case of the larger the better. At the margin the benefit of continuing to increase reserves is diminishing. So that I think right now we have large enough reserves already.”

China’s central bank buys dollars to prevent its currency from appreciating sharply, to keep its exports competitive. To do so, it sells yuan domestically and then issues local currency bonds to soak up any excess and prevent inflation. The cost to the central bank of this sterilization is the difference between what it earns on its reserves and what it pays to its bondholders.

In a world of negative real yields on US Treasuries, that cost is significant – not just to China, but to any foreign central bank loading up on the dollar. “They will get their principal back, but they have had to pay for the privilege,” says BlackRock’s Keeley.

But therein lies the opportunity, experts say: the global stock of foreign exchange reserves is excessive – and a percentage of such reserves could be used for more productive purposes at higher yields.

Infrastructure bonds

“There’s $11 trillion in reserves. That’s far more than what is required globally to maintain exchange rate stability,” says Keeley. “The reserves are not invested optimally. In the aggregate, too much is invested in government bonds that yield nothing and do nothing for global growth.”

Excess reserves could be put to more productive uses, such as investment in public goods like infrastructure, or in corporate debt that leads to private hiring, he says. “There’s an opportunity for central banks globally to invest for better macroeconomic outcomes, for example in infrastructure bonds. From a policy perspective, to have excess reserves languish in low yielding securities rather than solving the pressing problems of our times is unacceptable.”

Such an approach is being pioneered in Africa, for example, where the African Development Bank is poised to issue a $22 billion pan-African infrastructure bond in which the region’s central banks would invest 5% of their hard currency reserves.

The idea of central bank investment in infrastructure is one that Latin central bankers are willing to engage with – even if they remain publicly cautious about the practicalities of such moves.

Peru’s central bank governor Julio Velarde tells LatinFinance that while “it’s good to build more infrastructure and increase investment, you have to be conscious about the macro effects of too big an increase in domestic demand, including of course public expenditure.”

He adds: “Reserves have a counterpart which are liabilities. It’s not assets without liabilities. We have to consider that. Many times what’s not taken into account is that public investment tends to grow too fast which increases the prices of projects. That is happening around the world and many countries are still facing strong domestic demand.”

But central banks are nevertheless ideally placed to offer needed payments, custody and settlements infrastructure to allow best practice access to local bond markets for other central banks. In Latin America, says Mandeng, agreement on common standards, possibly within the framework of bilateral swap agreements, “would help to lower the barriers of entry for central banks”.

Brave new world

There is another argument in favor of greater reserve diversification: the evolving structure of the world economy itself. Experts say emerging markets’ growing share of global output should be represented by more diversified global reserve currency holdings.

The IMF forecasts emerging economies will comprise 48% of total global GDP by 2018. “That continuous shift in the real economy we have been observing for quite some time and which the crisis has actually brought forward – that still has no equivalent in the monetary sphere,” says Mandeng. “There’s still this fundamental asymmetry between reserve holdings and currencies and the shifts in the real economy.”

While US Treasuries offer advantages that emerging sovereign and currency markets can’t match in terms of depth and liquidity, Mandeng says: “we have to be realistic, central banks don’t need that much liquidity. They have shown in this crisis that they don’t need that liquidity and they’re not using those reserves.

“Emerging markets are becoming very important and yet if you look at their currencies, EM currencies are so far nowhere to be seen. It’s changing – but only slowly.”

Capital flows, trade and investment between emerging regions are greater than ever. But Mandeng acknowledges that both in terms of reserve currency diversification and investment strategy “one should not assume that central banks would be leading these trends. They are lagging a trend that we already observe.”

Not so fast

A shift in currency allocation and investment strategy by central banks – which would include channeling reserves towards an orderly expansion of local bonds markets – is proving more complicated than many, especially in the private sphere, would like.

Data on central bank reserve holdings and their composition is limited. But available IMF figures suggest that reserves and allocation policies remain, in the main, highly homogenous and concentrated in a narrow range of asset classes.

Central banks continue to pile into government securities, principally US Treasuries, as well as dollar and euro-denominated bonds issued by other advanced countries. The US dollar dominates strongly, with 62% of foreign exchange reserves denominated in the currency in 2012. This is followed by the euro with 24%, down from a 2009 high of 28%, and the yen and pound accounting for 8% of total reserves.

Although a risk-free rate for sovereign bonds post-crisis may no longer exist, for many central banks and institutional money managers, US Treasuries still remain the closest proxy for a safe asset.

This is especially true for Latin America’s central banks, which are widely considered to be among the most conservative anywhere. The case for reserve diversification by the region’s monetary authorities might be naturally constrained by their smaller size. On average reserves in the region account for some 13% of GDP, according to Moody’s. This contrasts with emerging Europe, where the median is 20% of GDP, and China, where they represent 40% of output; in Latin America, only Bolivia’s reserves ratio is comparable.

Still, the region’s reserve wealth continues to grow. Brazil now has $378 billion in foreign exchange reserves as its disposal. In Mexico, that sum is $166 billion, having expanded by 14% in 2012. Peru’s reserves increased by 24% in the year to February, to $67.6 billion. Colombia’s reserves are valued at $39 billion, the same amount as Chile’s. Uruguay has also continued to accumulate reserves, reaching $12.7 billion by the first quarter of this year.

Latin conservativism

But in general, the diversification of reserves away from US dollar securities has been slow across the region. “Latin central banks have been resisting that sort of change quite a bit. It’s changing, but you look at a region that is a laggard when it comes to currency diversification of reserves, with two or three exceptions,” says Mandeng.

Peru and Chile are widely acknowledged as being among the most innovative of the Latin reserve managers. Peru’s Velarde says that when it comes to reserves diversification, “the problem is that all assets are expensive now. Probably if we had diversified more before the crisis [that would have been preferable],” he says.

“Before Lehman we were already diversifying our portfolio. After Lehman, there was an increase in the purchase of dollars. Now we have started to diversify again, for example into commodity currencies. For example, the Australian dollar: we had very little before now we have an important quantity.”

Velarde points out that many central banks have legal restrictions on where they can put their money. “In our case, we cannot buy corporate bonds, for example. But we have been increasing the number of currencies we buy, public bonds, public agencies, multilaterals.”

Nevertheless, he says, the case for further diversification is tempered by uncertainty over the dollar. “The dollar has not been doing so badly in the last two years compared to other currencies. In spite of what is said about the weak dollar, we see the dollar actually as having been strong. In 2012 it was a little bit slow but still appreciating. And it has been strong this year.”

Similarly candid about diversification efforts is Chile’s central bank governor, Rodrigo Vergara, who tells LatinFinance: “We have already done that – we have already diversified our reserves.” The bank has shifted strategically into Canadian and New Zealand dollars, Swiss Francs and Chinese renminbi, but still has 40% to 50% of its reserves in US dollars, Vergara says. Still, he says: “we’ve had a very conservative management.”

Yet at the same time, the big central banks, led by Brazil and Mexico, remain reluctant to move.

Indeed, recent moves towards diversification have ended in tears for many. Having loaded up on gold in the years following Lehman, central banks – including those of Mexico, Brazil, Colombia, Venezuela and Paraguay – lost a combined $560 billion following the recent collapse in prices of the metal.

Says John Nugée, head of official institutions at State Street Global Advisors: “I don’t know what the effects of the current shake out is going to have on official sentiment. But people have been moving out, it does appear to be a genuine change in sentiment towards the metal.”

Diversification setback

Since the crisis, the most important source of reserve growth – global imbalances – has been shrinking as the US deficit and China’s surplus, respectively, grow smaller. A rebalancing of China’s growth model to prioritize domestic demand over exports will slow reserve expansion in the world’s biggest foreign exchange accumulator.

“We will make the trade surplus of China smaller, although we may still have a trade surplus but we will have a very small trade surplus as a percentage of GDP,” Yi says. “So we pursue a balanced balance-of-payments situation. In that case, we won’t have a rapid accumulation of reserves.”

Meanwhile, reserve growth in many instances – especially in Latin America – is now being driven by emerging countries buying dollars to combat rapidly appreciating currencies that accompany massive capital inflows.

Arnab Das, head of research and strategy at Roubini Global Economics says the implication of moderating global imbalances is a tendency not to diversify reserve currencies.

“It’s less of an investment strategy or return optimization strategy by central banks and more of an intervention-oriented strategy, which requires liquidity and currency matching. You want that currency diversification and that reserve pool at the central bank to match the original source of capital inflow,” he says. “All of those things point to reasons for less currency diversification.”

Indeed, despite the rhetoric, the process of reserve diversification has in recent years followed an extended rally in core bond markets, says Jan Dehn, co-head of research at Ashmore Investment Management.

“The crisis has been a setback for reserve diversification,” he says. A US Treasury rally, a deterioration in the risk/return of emerging markets currencies and elevated levels of uncertainty, among other factors, have “significantly slowed” diversification away from the dollar.

Dollar doubts

The more profound worry today, however, is over the soundness of prioritizing the US dollar at a time of great uncertainty about the trajectory of the economy and policy.

Much of that concern is about the likely path of interest rates once US economic activity picks up. But analysts are today increasingly split on how that will unfold.

A view is taking hold in the mainstream debate that conditions are ripe for another dollar rally: the deficit is shrinking as a proportion of GDP, employment is picking up and the housing sector has stabilized. Meanwhile, risk appetite is slowly improving.

As Velarde points out, the dollar has gained about 4% since the start of the year, on a trade-weighted basis, in line with riskier assets such as equities. The euro, sterling and the Japanese yen have retreated.

But another, more alarming scenario could be underway. Ashmore’s Dehn says it is implausible that policy could shift from today’s ultra-low interest rates and slack growth to higher rates and output “without having to go through a major macroeconomic adjustment”. This is especially so given the trillions of dollars the Fed has pumped into the banking system since the 2008 crisis.

“We’re only a couple of years away from the point where US household deleveraging ends and consumption rises,” says Dehn, who predicts a significant pickup in US domestic demand by the first quarter of 2016. “When that happens, inflation will rise too…we have the potential for quite a significant explosive evolution in inflation given where monetary policy is today, should demand come back.”

His argument is that the US Federal Reserve will face a choice between allowing a sharp rise in interest rates to their long-term average of 6.5% or engineering a gradual rise in rates. The former would trigger a “bloodbath” in the bond markets and lead to another financial crisis, whereas the latter entails a slow exit from quantitative easing – and an associated decline in the dollar of “at least 25%”.

“If 10-year Treasuries went to 6.5%, it would wipe out 35% of global fixed-income,” he says. “Given the enormous negative wealth effects, the Fed won’t allow it. They will unwind QE extremely slowly.” That will keep real yields negative.

Dehn says this scenario challenges the received wisdom of many central bankers. “It confronts them with the possibility – I would say likelihood – that their existing allocation is introducing risk into their reserves. This is something that will be experienced by all central banks that are heavily invested in Treasuries.

“This is beginning to appear on the radar screens of many reserve managers, who are now much more concerned about whether they have the right currency allocation in their portfolios.”

Dehn is not alone in anticipating a slow exit from QE. While he declines to comment on the likely direction of the dollar, SAFE’s Yi says that despite “a lot of market anticipation” and what he sees as “minor voices” within the Fed arguing for an earlier exit, extraordinary monetary stimulus will not be removed soon.

“I think [Fed chairman] Ben Bernanke’s and [vice chair] Janet Yellen’s message is very clear,” he says. “The US dollar is still the main reserve currency. I would like to see a stable dollar.”

The only way is up

BlackRock’s Keeley says reserve managers must now “be vigilant of the hidden fractures in the global economy. These could easily lead to further stresses on sovereign debt markets that have long been the mainstay of the reserve management world.”

The solution? Diversify. For Dehn, that means central banks should invest in emerging market local currency sovereign bonds. Why? Because the result of dollar weakness will be a rally in all major local currencies – in Latin America, the Brazilian real and Mexican peso, in particular. The challenge will be coordinating this diversification so as not to trigger a sudden realignment of global currencies, but the underlying trend is inexorable, he says.

There is compelling evidence, both anecdotal and factual, that central banks are now shifting away from the dollar and towards a broader portfolio of currencies and assets. And if Dehn is right, and if new reserve currencies are adopted wholeheartedly, the local bond markets could be on the cusp of precisely the boost so many would like to see.

But as Chile’s Vergara points out: “It’s a lengthy process, we don’t know other [non-core investment and currency] markets that well.”

He adds: “There’s a lot of learn in the process of diversification.” LF