Towards an inflation-target regime
By Lucas Llach, Deputy Governor of the Central Bank of the Argentina.
In 1582, western civilization began to use the Gregorian calendar, with a smart system of leap years (a leap year every four years, except for every 100 years, except for every 400 years: so that 1800 and 1900 were not leap years but 2000 was a leap year). This was a successful invention to measure dates that was globalized and it is used to this day. Almost two centuries later the humanity worked out the measurement problem: determination of longitudinal location in any part of the globe, much harder to determine than latitude. The Yorkshire carpenter John Harrison won, after 30 years of work, the award instituted by the British Crown in 1714. There is another more popular measurement method, the dimensions method that is also used nowadays: the decimal metric system. The marquis de Condorcet, son of the French Revolution, defined this method in 1971 as a product ´for all people and for all times´. Indeed, two centuries later this is the official pesos and measures system in the whole world, except for United States, Liberia and Myanmar.
These three elements (date, location, and dimension) are necessary to precisely define a sales contract of a country property: location, measures, signature of the contract and payment dates. There is only a quantitative element missing in a contract of this type: amounts payable. To define a clear value unit has been problematic for humanity. For centuries, the value of some metal was used as a standard. This system had problems since whenever the gold price rose, for example, in connection with the price of other goods; the result was deflation, i.e., fewer coins for the same shoes. Paper money enabled the separation of the nominal price of things from the relative price of metals and goods; however, paper money also caused inflation and, at times, deflation. Only in the past 30 years most of the countries in the world managed to achieve a relatively stable value pattern and, above all, predictable. For Argentina the problem of achieving a foreseeable and stable unit value was harder than for other societies. Argentina was almost always an inflationary country. Even worse, still this is an unresolved problem.
The reason behind inflation i.e. depreciation of the coin, changed in our country. Before the existence of a national currency there were provincial currencies, the most important was the Buenos Aires currency. In times of civil war and interprovincial conflicts, prior to the existence of the first national currency, military expenditure exceeded tax revenue. In 1881, following the defeat of Buenos Aires by Avellaneda and Roca, when Buenos Aires currency was replaced by the national currency, the former had already lost 96% of its original value.
In 1883, the first peso ´national currency´ completed the replacement of provincial currencies. Such paper currency circulated in one to one parity with a metal currency, the gold peso, as recommended by the manuals of the gold standard age. Such quality standard of maintaining a fixed parity between paper and gold did not last long: the first devaluation took place in January 1885. The gold standard lasted a year and a half. The expansionary policy of Roca and his successor Juárez Celman, which also required financing other provinces with scarce resources, concluded with the ´wallpapering´ (monetary financing to the national tax authority) and, ultimately, with the 1890 explosion. At that time, in the middle of the monetary instability, Juan Balestra wrote a book on that crisis: ´When in a time, surely distant, a pattern of invariable value has been found ´who doubts that we will look at this times with the same surprise that today we remember the times when we measured space with a foot or hand, the time with a watch sand and fever by the heat or cold felt by a doctor´s hand!´. And shortly after, still with the effects of the crisis, a British, Lawson quoted: ´Argentine people alter their currency as often as their change of president (…) No people in the world has an interest so deep in monetary experiments´.
After several decades of greater stability, during the second half of the 20th century, Argentine people lost their trust in the Argentine peso as a reliable unit of measure. No country in the world had so many years of inflation over 20% as Argentina in the last seven decades. As in the 19th century, part of the difficulty in providing the society with a stable and predictable measurement unit was that the currency production was used as a source for tax revenue. Nonetheless, at times inflation was not only a collateral effect of the Central Bank financing the government, but also a specific and deliberate target of the monetary authority: to undervalue the currency, depreciate it. This was a strategy to undervalue wages measured in international currency, a competitiveness shortcut that did not last in time and lead to the inflation we have nowadays. At other times, the anxiety to find a stable measure value led us to a different short cut: link our currency value with foreign currency in order to, so to speak, enjoy their stability. Nonetheless, stuck our currency value to another country currency in order to have a foreseeable and moderate inflation implied that the value of our goods and services changed similarly to that of the standard currency. So, in a changing world that method finally breaks. The result of currency fixations was an inflationary explosion (generally sooner than later) after the government or the market were persuaded that argentine prices in international currency were not sustainable at such exchange rate.
In other words: Argentine currency was never a foreseeable and stable value unit because the monetary policy always had other targets: deficit financing or keeping an undervalued currency or - as an anti-inflationary shortcut that confused means with ends – set Argentine currency value with regard to other currencies.
During 2016, the Central Bank of Argentina changed to the inflation target regime, which departs from an obvious premise: if we want to control inflation, the Central Bank´s target must be to control inflation. Further, I make a game of seven questions and answers, among others, on the inflation target regime:
1. Is the inflation target contradictory to other Central Bank´s targets? In accordance with the Charter of the Central Bank, the purpose of the Central Bank is to promote financial stability, employment, and economic development with social equality. It is quite obvious that foreseeable and moderated inflation favored financial stability, a necessary condition for a sustainable economic development and that limiting the inflationary tax favored social equity. However, is there a conflict between price stability and high levels of employment?
The well-known ´ Phillips Curve´, which is an empirical regularity originally described by the north American economist Irving Fisher in 1926, postulates a negative relationship between inflation and unemployment i.e. more inflation, less unemployment and vice versa.
However, nearly half a century later we are aware that he Phillips Curve has a trick. There might be a negative effect in employment – explains the consensus of economists – if the monetary authority induces an unexpected disinflation. For example: if in a particular year contracts were signed with an inflation expectation of 10% but the monetary authority seeks a 10% inflation, the mismatch between the expected and the realized outcome may be recessive as it may generate cost increases higher than price rises. Therefore, unexpected disinflation may be recessive for a while and, conversely, unexpected inflation may be expansive for a shorter period. However, a simple relationship between inflation and unemployment cannot be established, in fact, to set an example during the last 15 years the three years with highest inflation were also years of recession (2002, 2014 and 2016).
The inflation target regime precisely seeks completely the opposite to inflationary or deflationary surprises. This regime seeks foreseeability: to announce an inflation target or range and to use instruments of monetary policy with a view to fulfilling its targets. To the extent that such range or target is trustworthy and the economic players incorporate such expectation; there is nothing better for economic activity and employment than to validate those expectations with the monetary policy.
For the same reason, the inflation target policy does not imply that the monetary policy countercyclical role should be abandoned. When a country with 5% inflation target suffers from recession, the expected impact of a lower level of activity should be a drop in inflation below such target. In order to prevent inflation from dropping below the level sought, monetary policy has to be expansive in that context, precisely in that moment the economy needs to avoid unemployment. In every case, credibility is a fundamental factor in a viable target inflation scheme. This leads us to the second question.
2. Why is there target gradualism? During this year of transition to inflation targets, the Central Bank seeks 1.5% monthly inflation or lees during the last quarter of the year. The last Market Expectations Survey carried out at the end of August revealed that, leaving aside the expected effects tariff changes (disinflationary in August and September, plus the expected inflation in some of the following months), experts forecast 1.6%, 1.5% and 1.5% for nucleus inflation in October, November and December. That is to say, Central Bank´s inflation goal is close to be completely credible for the remainder of 2016. Regarding next year, inflation expectations are placed at 19.8% and in the case of inflation nucleus at 17.9%. As the target announced by the Central Bank is between 12% and 17% for 2017, we still have a so called ´credibility gap´ of almost 3 percentage points in connection with 17% cap. The Central Bank is leading a monetary policy to eliminate such credibility gap.
We believe that target gradualism is important to achieve the credibility goal. Our daily duty is to convince economic players that the monetary policy will adjust along the following months in order to achieve an inflation lower than 17%, as we convinced the society that the monetary policy will adjust to achieve 1.5% goal or lower in the last quarter. This is a possible goal and to set a more ambitious goal would be much harder. Therefore, we adopted a gradual path to disinflation that includes 5% goal as from 2019. This leads us to the third question.
3. Why 5% and not less as final target? At this stage many central banks of developed countries with 2% inflation targets are wondering if this target may complicate monetary policy for being too low. In fact, many countries face the deflation trap and cannot be unwind by negative interest rates (as in many countries), in order to return to the path of economic expansion and small but positive inflation for their target. At this time, in the world countries discuss if a higher target can grant a greater margin for maneuver with a 5% inflation target, for example 0% interest rate (almost a floor for the monetary policy) is more expansionary with 2% inflation as it implies a lower real interest rate.
For an emerging country like Argentina, frequently exposed to external shocks 5% target may be appropriate for other reason. When the economy needs to become cheaper facing the world (for example by effects of the international crisis) 5% target allows accommodating such impact with a monetary depreciation without increasing inflation or decreasing deflation. If instead there was 0% inflation target, for example, an adjustment of relative prices towards a greater real exchange rate may imply nominal deflation of certain prices so as not to deviate from the target. We know that even though we do not believe in the Phillips curve, price deflation may require a recession.
Once again: the target inflation scheme with gradualism and reasonable targets is not contradictory with full employment, quite the opposite.
4. Can the peso act as standard measure if there is disinflation for more than three years and yet only reaches 5%? With one digit inflation, prices begin to make sense. During high inflation periods, nobody knows the price of things (with this in mind, marketing experts cause supermarkets to compete by advertising discounts instead of prices). Regarding the comparison between current and future prices, to the extent that the Central Bank gradually fulfills its targets and the economic players align their inflation expectations to the announced targets, contracts may incorporate such inflation expectations more precisely. Financial contracts shall have decreasing nominal interest rates.
Therefore, the Central Bank believes that in order to deal with this disinflation process and, particularly thinking in the long term, allowing the signature of inflation protected contracts makes sense, even in the case of 5% inflation. Consequently, following the successful method of Chile, Colombia and Uruguay, the Central Bank promoted an account unit protected from inflation that follows consumer price index. One of the advantages of the account unit protected from inflation is that it allows associating over time, loans installments to economy nominal payment capacity. Even with 5% inflation, a set nominal payment executed within 25 years has 70% less real value if that same nominal payment would be executed today. Therefore, in a fixed installment loan the initial installments should bear the value lost in the final installments. This sets an access barrier to long term credit by low income families that cannot face high initial installments. Today such barrier is huge. The unit account system has a real constant value that allows associating each payment to the level of prices in force at any time.
5. Does the inflation target scheme imply higher interest rates? On the contrary, in the long term anything that implies greater predictability implies less risk and, therefore, less extra rate connected with that risk. Interest nominal rates, in regular conditions, must reflect the expected inflation, usually when disinflation is sought, a real positive rate and an extra real rate. It should be noted that monetary policy rate currently set by the Central Bank (Lebac with 35 days maturity) is a short term rate. In order to determine the bias of monetary policy, nominal interest rate must be compared to the corresponding period. For example: while the Central Bank kept a 38% interest rate between March and May, the equivalent in monthly inflation rate was 3.16% in comparison to the expected inflation for those months.
The last two questions are connected with objects capable of awakening passions and that I barely mentioned so far: amount of money and exchange rate.
6. How does the amount of money play in the target inflation scheme? In a scheme where the monetary authority sets the interest rate, the amount of money (whether ´monetary base´ or other aggregates) is a consequence, a result, an institution that has no arrow pointing anywhere. As they say in the central bankers jargon: ´we did not abandon monetary aggregates, they abandoned us´.
How does this system exactly work? Let’s analyze the case of Argentina today. Each Wednesday the Central Bank sets the interest rate for the sale of its securities with 35 days maturity. At the same time (and more important) such rate shall intervene the following days of the week in the securities market. Let´s think about that market from banks or other financial agents’ point of view. For the sake of simplicity, we can imagine that they have to choose between 3 low risk assets: liquidity in pesos with no interest payment, Lebac with interest payment or foreign currency. If for whatever reason the demand for liquidity increases, securities may automatically be sold at a pre-determined rate and the Central Bank shall buy those securities. In contrast: if there is surplus liquidity, they may buy as many securities as they may need to in order to get rid of such surplus liquidity and the Central Bank will be prepared to sell those securities (i.e. absorb liquidity) at the monetary policy rate. In other words, the demand for pesos determines the monetary offer and not vice versa. Indeed, by controlling the interest rate the Central Bank may influence the amount of money demanded (the lowest the rate, the less expensive is to keep liquidity); however the look of the Central Bank is on the rate.
In order to think of the monetary policy effect on inflation is not necessary to analyze monetary aggregates. The interest rate affects the inflation course by channels that are independent from the amount of money. In the first place, by altering expenditure and saving incentives. Secondly, because the monetary policy affects the domestic currency price compared to other currencies even though this is not an intentional goal.
Thirdly, but more important, by its value as a commitment sign: provided that the public observes that the monetary authority employs the interest rate to obtain inflation targets, includes such targets in its expectations and such inflation expectations influence inflation as they are included in contracts (financial, rent, employment, etc.).
Frankly, we believe that the Central Bank had a successful year in convincing those that closely follow the Central Bank actions that the monetary policy reacts in order to achieve its inflation targets. Every new inflation datum is read by market experts as a sign of how the Central Bank will act. The belief that the Central Bank is being responsive to inflation is a necessary condition for the inflation target regime. In addition, the belief that the Bank´s policy is the most appropriate to fulfill its goal, something we are close to, is already a sufficient condition. After all, a possible definition of the inflation target regime is a regime whereby the society believes in the monetary policy implemented that will lead to the inflation set in the targets.
7. And the dollar? My closest friends know they cannot do it but anyway they ask me: should I buy dollars? or is the rate convenient? Even though I knew the answer, I would answer I do not know of course; however I am pleased to know I actually do not know the answer. Maybe the greatest virtue of the inflation target regime with floating exchange rate is that nominal currency depreciations or appreciations are never anticipated and, therefore, never generate an obvious speculation towards one side or the other. The reason is simple: if everyone knew the currency will depreciate, it will depreciate immediately, and people would stop thinking that it is obvious that it will depreciate because depreciation would have already taken place. Something similar might be said in connection with appreciation.
Therefore, what cannot be anticipated on the nominal exchange rate, neither can be said in regard to the real exchange rate. Even if economic science discuss the empiric validity of the so called ´discovered parity of the interest rate´, the truth is that economists in general believe that there must be a certain relationship between the interest rate in pesos (the amount paid by a Lebac) and the expected devaluation (in terms of pesos, the amount obtained by staying on dollars). If that was the case one of those assets shall not be requested.
However, let´s combine that matching idea between interest rate and expected inflation with, for example, a monetary policy that intends to set a real interest rate slightly positive. Such slightly positive real rate means that the nominal interest rate is higher than the expected inflation. But then, in expected value, a slightly positive real interest rate policy implies that the market is waiting for a nominal depreciation barely higher than the inflation rate or in other words a small increase in the real exchange rate. Of course, such expectation is in fact something like the expected value in a probability distribution. In this sense, markets believe that the real exchange rate may rise, drop or remain constant.
Am I saying anything on foreign exchange policy? No. I am only saying that if people should believe in financial market prices, the following issues cannot be said with certainty: that the exchange rate shall move towards one side or the other as indicated by the expected depreciation curves; that the real exchange rate shall obviously go in one direction or another.
Are there many uncertainties? I do not think so. I believe there are the appropriate ones. Under an inflation target regime the monetary authority does not seek for certainties.
Under an inflation target regime the monetary authority does not seek certainties on subjects that cannot give certainties, as they do not depend on the Central Bank action, it is incapable of giving certainties. The inflation target regime seeks to be as foreseeable as possible on the subjects that depend on the monetary authority: the purchase power of that currency, whose care was entrusted with.
Our everyday duty is that the Central Bank lives up to that responsibility. Thank you.
September 8th, 2016.