William Davidson Institute Working Paper 402
shock to output (favourable or not) pushes the system on a circular trajectory, and
brings about costly fluctuations. A very powerful shock may even push the system on
a very peripheral trajectory; in this case, there may be a significant risk that such a
trajectory is not consistent with the condition R>0 and the currency board cannot
survive. Of course, the dynamics put forward in this simple model depend on the
specific form used to describe price and output formation. Other expressions may lead
to more complex dynamics, where main variables follow explosive or chaotic paths.
Finally, even under a currency board there is a money multiplier11 and thus a
risk of generalised run on banks. Furthermore, this risk is somehow higher than with a
normal central bank, given that the currency board has little possibility to act as a
lender of last resort (Berg and Borensztein, 2000; Guide, Kahkonen and Keller, 2000).
As argued by Chang and Velasco (2000), currency boards make balance-of-payment
crises less likely only at the cost of making bank crises more likely. This should be
seriously considered by policymakers in those countries that have fragile banking
systems - as is the case of transition economies. And as put forward by Argentinean
experience, even a country endowed with a currency board may face major liquidity
risks, where foreign investors may refuse to roll over outstanding debt. In this case,
the government may be forced to default, even if the long-run condition for solvency
is satisfied (i.e. total debt is lower than the discounted value of future primary
surpluses).
4.3 Self-fulfilling prophecies and speculative attacks
As the French experience of the early nineties has shown, even strong
currencies may be subject to speculative attacks. Such logic would also justify attacks
on currencies backed by currency boards. Standard explanations (Obstfeld 1984,
1992) develop on the logic of self-fulfilling prophecies (multiple equilibria).
Basically, the decision to attack a currency is motivated by a large expected gain, that
will materialise if the attack is actually undertaken. The scope for large expected gains
results from the strong asymmetry between ex-post gains and losses of the
speculators. Consider a speculator who borrows 1mld lei for one month and buys
euros, while simultaneously he sells euros and buys lei in the one-month forward
market. If at the end of that month the parity is unchanged, he will bear the transaction
costs (and the interest rate), let say 1% of the principal. But if at the end of the month
the leu has depreciated by 6%, the speculator will have made a 5% gain, an equivalent
of a 60% yearly interest rate. Even if the probability of the devaluation is of 0.5, the
domestic interest rate should increase by 30% in order to counter the attack. If the
business sector cannot resist to this shock, devaluation is the only choice.
Of course, the action of one isolated speculator could not destabilise the
currency. But what is optimal for one speculator should be optimal for many others,
depending on their specific characteristics like the information set, wealth, and so on.
Chances that a run on the currency occurs are more significant if the country is
globally perceived to be in a position where it cannot sustain the parity. For instance,
11 Let us assume that commercial banks hold a proportion φ of their deposits in the form of notes
(themselves covered by dollars within the currency board): NB=φD. (For simplicity, we neglect here
the banks deposits with the board). The public would also hold banknotes as a proportion b of their
deposits, NP=bD. The money stock is M=N+D=NB+NP+D=(1+φ+b)D. Then, the ratio between the
money stock and notes is m=M∕N=(1+φ+b)∕(φ+b) higher than 1. Thus, there is a strict relationship
between the stock of foreign currency and the money stock. The smaller φ and b, the higher the
multiplier is.