The channels of transmission of monetary policy∗
Juan Antonio Morales
President of the Central Bank of Bolivia
I would like to comment on the difficulties of establishing a domestic monetary policy in
countries in transition from the nominal anchor of exchange rate targeting to inflation
targeting cum flexible exchange rates. Before reaching full-fledged inflation targeting, the
channels of monetary policy, it will be argued, are likely to be weak. A related but
unanswered question concerns the appropriate timing to move to full-fledged inflation
targeting, and hence the length and characteristics of the transition period.
In countries where there is a large degree of de facto dollarization, which implies to a
significant extent a hard peg, there is little connection between monetary policy in domestic
currency and financial markets. This lack of connection is reflected, inter alia, in very
volatile money market rates in domestic currency, very distant from the rates in dollars, the
prevailing currency in financial markets. Arbitrage conditions between the two currencies
are seldom met and the interest rate spreads tend to be very volatile.
After abandoning the exchange rate anchor, monetary independence has to recover and a
whole new monetary policy has to be constructed, including the guidance of expectations
by a credible central bank. Institutional development is not an easy matter and often the
Comentarios expuestos en el Segundo Seminario de Alto Nivel del Eurosistema y de los Bancos Centrales
Latinoamericanos. Río de Janeiro, 26 de Noviembre de 2004.
legal support for central banking is inadequate, despite the proclaimed independence. In
addition, financial markets are shallow, with underdeveloped capital markets and with most
of the financing of domestic investment granted by banks. Recurrent fiscal weaknesses
cause large placements of public debt and banks’ portfolios are overloaded with
government paper. It is not hard to see that the channels of transmission of monetary policy
are going to be weak, when they are not entirely inexistent.
Especially in countries that in the past have experienced high inflation, dollarization has
been part and parcel of the reconstruction of their financial systems. This is an important
point to keep in mind: financial development has been closely related to dollarization.
Again, with financial development highly anchored in a foreign currency (and the attendant
foreign institutions) the conventional channels of monetary policy are either weak or work
in unexpected ways with regard to conventional theory.
My comments are very much based on my Bolivian experience, although Bolivia is by no
means a typical Latin American economy. She is a poor counterpart of the more mature
emerging economies in the region. Yet, for analysis, Bolivia provides and extreme
illustration of the roadblocks that plague many countries in their way from exchange rate
anchors to other types of nominal anchors. As important, Bolivia has escaped financial
collapse, despite her lack of monetary independence, her intermediate exchange rate regime
and very difficult fiscal and political conditions
An independent monetary policy is needed to give sense to the transmission mechanisms.
In turn the conditions for an independent monetary policy can be very demanding. If we
focus specifically on the traditional channels of transmission of monetary policy:
- the interest rate channel
- the exchange rate channel
- the credit channel
- the expectations channel,
we can see why they work languidly:
The interest rate channel. Given the difficulties of establishing the term structure of interest
rates, the monetary policy interest rate has a very limited reach, at best to the interbank
market. The weak transmission originates in the shallowness of financial markets and the
uncertainties in the inflation rate, which can be low but its variance may be still high. Also,
the volatility of inflation makes difficult to gauge the real interest rate.
In addition, financial uncertainty is usually high and the bond markets are rather illiquid,
leading to high and variable risk premia. In particular, in Bolivia, bank lending rates seem
determined mainly by risk premia. The expectations mechanism in the determination of the
term structure of interest rates cannot work in this context.
Also the fact that inflation is not the main focus of attention of financial markets has a
bearing on the formation of the yield curve. In the developed financial markets of industrial
countries, it is well known that with lower expectations of inflation and hence lower
expected future interest rates, long term interest rates would be lower today. However, even
in mature emerging markets, the transmission of changes in interest rates takes a relatively
long and variable amount of time to materialize across the economy.
It is very unlikely that the policy interest rate would affect in a significant way the interest
rate of long term bonds, or the financing of capital goods or even of housing. The problem
is compounded when de facto dollarization is widespread, where the term structure of
deposit and bond interest rates is mainly determined by the U.S. market rather than in the
domestic market. However, in econometric studies in my country it has been found that the
interest rate of the 13-week Treasury bill in dollars Granger causes other longer-term
interest rates in foreign currency, including the interbank rate.
The exchange rate channel. Also this channel works weakly in an emerging economy in
transition from exchange rate anchoring to inflation targeting cum flexible exchange rates.
It is well known that the exchange rate, being an important transmission mechanism, will
normally enter in the policy evaluation models of small open economies. Exchange rate
movements passes-through to the inflation rate, requiring thus strong reactions of the
interest rate in the policy function, which may unduly affect financial markets. Policy
reactions to exchange rate shocks are typically strong, and in inflation targeting economies,
the response is usually stronger than to the inflation rate and the output gap. Also exchange
rate movements have a direct bearing on the solvency of the banks, if there is liability
dollarization. Given these conditions, the monetary authority will try to keep exchange rate
movements within narrow limits. There will be a pervasive “fear of floating”.
Ironically and notwithstanding what has been said, the exchange rate has become a main
monetary instrument (that can be a close substitute of a monetary policy rule) in my
country. With the exchange rate we follow a special “Taylor rule”, in the sense that our
monetary rule is exchange rate based, rather than interest-rate based. In the Bolivian
incomplete crawling rate system the rate of crawl adjusts to gaps in inflation, economic
activity and the real exchange rate.
The credit or lending channel. It is either narrow or when it works it does somewhat
unconventionally. Banks are the most important players in the financial systems of most
Latin American countries. Their loans are very imperfect substitutes of bonds as source of
finance and moreover banks are unable to insulate their supply of credit from reductions in
deposits, induced in part by changes in monetary policy. One could expect then a very
strong lending channel that gives extra push to the effects of monetary policy on aggregate
demand. Yet, this may not be the case because expansions and contractions of domestic
money have a limited impact when most of the lending is in dollars. Also, dollar deposits
are very volatile and prone to runs when economic or political conditions deteriorate.
Of course, there is the possibility of conducting a monetary policy in dollars, as we do in
Bolivia, but this unusual. There is also the intriguing possibility to consider three types of
assets in the banks’ portfolio: loans in dollars, loans in domestic currency and bonds. Prima
facie, these assets are imperfect substitutes and hence monetary movements in domestic
currency may have, after all, consequences both in the dollar lending of banks and the
overall size of credit by banks.
The balance sheet effects of the lending channel may also be weak or perverse. Monetary
policy in domestic currency per se hardly affects the price of assets, either financial or real
like housing, which in many Latin American countries are priced in dollars. No major
changes in creditworthiness may ensue changes in domestic money. However, monetary
policy, to the extent that it is conducive to exchange rate depreciations, while restraining
inflation, could have vast and unsettling balance sheet effects as the real price of assets
denominated in dollars will increase as well as of dollar denominated liabilities. With a
highly leveraged private sector, moreover with dollar debts concentrated in non tradable
sectors, an expansionary monetary policy may end being contractionary and reducing the
level of lending.
On the other hand, the net wealth of depositors in the banks may increase by the amount of
the real depreciation hence real cash balances increase and work their way through
conventional IS-LM effects.
The expectations channel. After leaving the exchange rate as a monetary anchor, the
expectations channel has to be created from scratch. The public’s attention will continue to
stay on the exchange rate for a long time while inflation, unless it is very high, commands
relatively low interest. The central bank’s commitment to low inflation takes time build and
to be internalised by domestic financial markets and the public at large, given their
traditional focus on the exchange rate. The time required will be the longer the weaker the
fiscal situation and, for small open economies, if there are contagion effects from disorders
Without significant reductions in the levels of dollarization, the possibility of an
independent monetary policy, under the assumption that it is desirable, is small. The
manoeuvring space for monetary policy in domestic currency in the presence of
widespread de facto dollarization is very narrow and, as important, it is perceived by the
public as marginal. The main negative features of dollarization are first, that it leaves the
country essentially without a monetary policy and second, that it can be a source of
financial instability to the extent that there are currency mismatches.
Dollarization appears so resilient because of the weakness of fiscal, financial and monetary
institutions. While actual inflation has been low for many years, there are still in the public
expectations of abrupt exchange rate depreciations and ensuing high inflation. Payments
dollarization, which has a bearing on financial dollarization, has increased because of
The manifestations of an acute “peso problem” appear not only in the high spreads between
interest rates in domestic currency and dollars for most financial operations (although not
for all) but also in the reluctance of financial institutions to receive deposits or grant credits
in domestic currency.
Unless the credibility of fiscal and monetary institutions is greatly and lastly strengthened,
dollarization that implicitly carries imported credibility will persist. Laws that assure the
independence of central banks, a floating exchange rate (with all the risks it entails during
the transition) and market-friendly mechanisms like the introduction of bonds linked to
inflation and advantages in taxation and regulation for assets denominated in domestic
currency may help to revert dollarization, but the thrust has to be on confidence.
The bank lending channel can be a very powerful channel of transmission of monetary
policy enhancing the latter’s effect on aggregate demand, at least for a while, but provided
again that dollarization is reduced. Meanwhile, ways and means for smooth functioning of
the banking system have to be sought emphasizing prudential regulations and including
rules of internalization by the banks of the costs of dollarization.
Notwithstanding the limitations faced by the central banks of emerging economies, they
can still have non negligeable impacts on expectations, and ultimately on prices and the
level of activity, by following predictable rules and communicating very clearly both their
commitment to low inflation and their reaction function.