Capital flows are rarely discussed without a genuflection in the direction of the impossible trinity, also known as the trilemma. For example, Magud et al (2011) write: “… a trinity is always at work. It is not possible to have a fixed (or highly managed) exchange rate, monetary policy autonomy, and open capital markets.” According to the trilemma, a stable exchange rate without capital controls requires domestic and foreign interest rates to be equal. Otherwise, ‘uncovered interest arbitrage’ will force continuous appreciation or depreciation of the currency. As such, nations without capital controls must choose between stabilising the exchange rate (by slaving interest rates to foreign rates) and stabilising the domestic economy (adjusting interests slaved to domestic macro conditions but letting the exchange rate fluctuate). Mechanically, this is enforced – according to trilemma logic – by substantial capital inflows or outflows and the impact of these on the money supply.
Why this doesn’t fit the East Asia experience
Since the 1997–98 Asian crisis, East Asian countries have clearly run their own independent monetary policies.1 They have successfully set interest rates to broadly achieve their inflation objectives. As Figure 1 shows, they are most definitely not all slaving their rates to foreign rates.
Despite this, their exchange rates have been fairly stable. They have managed their primary exchange-rate objective – leaning against the prevailing appreciation pressures in order to maintain international competitiveness (see Figure 2). Remember that according to the classic trilemma, the similarity in exchange-rate movements since the global crisis should have coincided with identical interest rate levels (all equal to, eg, the US interest rate); comparing Figures 1 and 2, we see this isn’t the case.
These attempts to restrain appreciation have involved heavy government intervention, resulting in very large increases in foreign-exchange reserves (Figure 3). This didn’t, however, cause excessive increases in base money (Figure 4), thanks to effective sterilisation by open-market operations and increases in banks’ required reserves.
Figure 3. Foreign-exchange reserves as a share of GDP
Figure 4. Growth in foreign-exchange reserves (y-axis) and base money (x-axis), Percent, 2001–07
Why doesn’t the trinity apply?
There are four reasons why the trinity doesn’t work in East Asia.
First, if uncovered interest parity held, markets would treat different currencies as close substitutes. An investor would know that the interest differential would be a good guide to where the exchange rate was heading and even small interest differentials would trigger large arbitrage flows. It is now abundantly clear that interest parity offers feeble guidance for the exchange rate–interest rate nexus (see Engel 1996). The parity condition often gets the direction wrong, let alone the quantity (Cavalo 2006), as it does for six of the seven countries illustrated in Figure 5.
Figure 5. Annual average interest differential versus change in exchange rate 2001–10
Capital flows responding strongly to interest differentials are the core element in the impossible trinity story. But in practice:
- Different currencies are not close substitutes; and
- Capital flows are driven by many other forces besides short-term interest differentials.
Second, instead of well-formed views on how different currencies will behave over time, there are fluctuating (sometimes wildly fluctuating) assessments of risk attached to cross-currency holdings. The higher interest rates generally available in emerging countries have encouraged carry trade–type capital inflows, but these were offset by official reserve increases (Figure 6).
Figure 6. Net capital flows to emerging countries ($ trillion)
Third, the impossible trinity envisages that any intervention to prevent these capital flows from bidding up the exchange rate will be fully reflected in base money increases which will, in turn, thwart the authority’s attempts to set interest rates as desired. But this sort of base money-multiplier view of monetary policy no longer corresponds with the way monetary policy works in practice.
These days the authorities set the policy interest rate directly via announcement, while managing liquidity in the short-term money market through open-market operations, including an effective capacity to sterilise foreign-exchange intervention (Figure 4). In some cases (eg China) excess base money was effectively sterilised through increases in banks’ required reserves. Thus capital flows do not usually prevent the authorities from setting interest rates according to their objectives.
Finally, the impossible trinity envisages that any official intervention in foreign-exchange markets will be taking the exchange rate away from its equilibrium, opening up arbitrage opportunities. But suppose, instead, that the authorities have a better understanding (or longer-term view) of where the equilibrium lies, and are managing the exchange rate to maintain it in a band around the equilibrium. East Asian countries have not, in general, prevented some appreciation of their exchange rates, but they have sought, through intervention, to prevent momentum-driven overshooting.
Is there a useful softer version of the impossible trinity?
Even if the impossible trinity in its pure version does not hold, is it still a useful concept in a looser version, as a reminder that there are interconnections and policy constraints between interest rates, exchange rates, and capital flows?
Frankel  As they become more closely integrated internationally, foreign investors will increasingly respond to this underlying profitability differential. How can this prospect of sustained higher returns be reconciled with portfolio balance for the foreigners whose initial portfolios are in the lower-return mature economies? This, not the short-term impossible trinity problem, is the policy challenge
The impossible trinity began as a useful theoretical insight into the interactions of policy instruments. It is still a useful blackboard reminder that not all policy combinations are possible. The blackboard illustration, however, has been adopted as a doctrinal policy rule. This over-emphasis on a simple thought-experiment may have been because it served to support the arguments for free-floating exchange rates. The argument went like this: capital controls are not workable; if you want to have your own monetary policy, then you have to let your exchange rates float freely.
But the impossible trinity was a stylised insight relying on simplified assumptions. The real world was always more complex and nuanced. Of course there is some connection between interest differentials and capital flows. But there are other forces motivating capital flows, and these are much more random and non-optimising than envisaged by the impossible trinity. The fickle changes in risk assessments, mindless herding, and booms and busts in the capital-exporting countries make international capital flows volatile in ways not envisioned in the trinity.
Author’s Note: This column is based on ‘The Impossible Trinity and Capital Flows in East Asia’, Asian Development Bank Institute Working Paper 318 November 2011.
Aizenman, J, MD Chinn, and H Ito (2009), "Surfing the Waves of Globalisation: Asia and Financial Globalisation in the Context of the Trilemma", Asian Development Bank Working Papers No. 180.
Cavalo, M (2006), "Interest Rates, Carry Trades, and Exchange Rate Movements", FRBSF Economic Newsletter 2006/31.
Engel, C (1996), "The forward discount anomaly and the risk premium: a survey of recent evidence", Journal of Empirical Finance (32): 305–319.
Frankel, JA (1999), "No single currency regime is right for all countries or at all times", Princeton Essays in International Finance 215.
Magud, NE, CM Reinhart and KS Rogoff (2011), "Capital controls: myth and reality – a portfolio balance", Peterson Institute Working paper 11-7
1 Except, of course, Hong Kong, with its fixed rate. Singapore is a special case, implementing monetary policy via the exchange rate rather than interest rates. Its capital market is open; it closely manages its exchange rate; and it has an independent monetary policy, achieving its objective of having one of the lowest inflation rates in the world.
2 Some might see this same argument in terms of growth rates. Interest rates will approximate the economy’s growth rate (whether measured in real or nominal terms). Thus the higher prospective growth rates of the emerging countries will be accompanied by higher interest rates.