This page accompanies my presentation at the annual meeting of Business Iceland on April 7, 2016.
My slides for the event (in Icelandic) can be downloaded here, and the recording (in Icelandic) is here
The outline of the talk is:
Iceland has one of the most volatile economies in the OECD.
Over 1960 to 2014, economic growth in Iceland is about average for OECD, while macroeconomic volatility is the sixth highest.
The data is obtained from the World Bank World Development Indicators (WDI) . The specific variable used is NY.GDP.PCAP.KD.ZG, defined as “Annual percentage growth rate of GDP per capita based on constant local currency. Aggregates are based on constant 2005 U.S. dollars. GDP per capita is gross domestic product divided by midyear population.”
I filter this to only include OECD members. That raises some issues, should one only include data from the year a country joined the OECD, and then how to deal with the non-synchronous sample sizes. I just use all data available for current members.
The following figure shows average growth over the time period, and Iceland’s is exactly average, but is a bit higher than the median country.
There are many ways one can model macroeconomic stability. I opted for one of the simplest, the annual volatility of GDP growth.
This would not be too bad if the high volatility came hand-in-hand with high-growth, in a risk—return story. But, its not so. The plot below shows the mean-volatility relationship along with the regression curve:
Iceland has average return — high volatility, while most of northern Europe has low return — low volatility.
These results are not surprising given how Iceland gets its growth.
One explanation for the average growth — high volatility result is that GDP level effects are driven by particular factors,
Iceland is particularly prone to boom and bust cycles.
While Iceland has created a number of high-tech companies, industrial developments follow the typical pattern of companies migrating to high-tech clusters, the main exception is in fishing technology, where Iceland might be a cluster. While the economic geography experts argue that it is difficult for any region to stand against such forces, the macroeconomic instability in Iceland certainly does not help.
Iceland has the lowest return on education in Europe.
There is no single correct way to get returns on education. While I am sure there are many more sophisticated ways to do it than the below, this is at least approximately correct.
Eurostat publishes data on “Mean and median income by educational attainment level (source: SILC)”, download here. Pick two variables “ED0-2 Less than primary, primary and lower secondary education (levels 0-2)” and “ED5-8 Tertiary education (levels 5-8)” and plot how much larger the second is. Data is for 2014.
This figure however overstates the return on education. A person who finishes tertiary education has spent perhaps eight years more in school than a person who leaves school as soon as possible. So, the latter is earning money for eight years while the former is earning a lot less, on average.
So, if one does a back of the envelope calculation, discounting future earnings to the age of 15, assuming that everybody works to the age of 70, and that it takes eight years to get tertiary education, where the discount factor is and salary is .
Define the NPV of the return on no education as and tertiary education as
Then, it is quite easy to identify whether education increases or decreases one’s lifetime income.
Economic activity is driven by return on natural resources not human capital.
One manifestation of this is the low return on human capital. If the economy is driven by the exploitation of natural resources, it is the return on these very same natural resources that drive economic activity, profits and salaries and not human capital.
And since the main drivers of growth are based on relatively unskilled labor, there is no wonder that the return on education is low.
In turn, the resulting commoditized educated labor force, especially those without an international market, will be low-paid. This explains the flat live cycle profile of salaries.
If the authorities want to create a better environment for future economic growth, and especially for high-tech companies based on human capital, they need to create stability.
There are several sectors one can look at
The labor market to in Iceland is highly procyclical.
A key reason is centralized bargaining and the emphasis on maintaining relative parity and small differences between the compensation of the various categories of workers.
Historically, in up-cycles, often because of level effects in natural resources, salaries get bid sharply up across the board. This then causes increasing difficulties for those sectors not benefiting from the level effects.
In the down cycle, salaries are too high, culminating in high inflation working as an adjustment mechanism.
This does not create undue difficulties for firms engaged in those natural resource sectors not subject to international competition for labor, or firms that are essentially domestic and not subject to most international competition. However, other firms that operate in the international domain will suffer.
Establish a sovereign wealth fund.
Fiscal policy has been countercyclical in recent years, with the central government maintaining a large surplus in upswings.
In order to get more stability, the central government could go one step further and establish a sovereign wealth fund, invested abroad.
In years when economic growth is positive, the central government invests in the sovereign wealth fund, drawing on it in recessions.
This would help both with reducing business cycles and the execution of monetary policy. On the latter, when the central bank does unsterilized interventions, the effect is partly offset by the sovereign wealth fund investments, while the return on the sovereign wealth fund also partially offsets the cost of the intervention.
The government is now in the unusual position of owning most of the banking system. In turn, these banks also own a sizable part of the remainder of the financial system.
When it comes to privatization, the government has two choices:
The government can either maximize the revenue from the privatization or maximize future economic growth. These two objectives are mutually exclusive.
The financial system in Iceland is highly oligopolistic, inevitable giving the small size of the economy. This implies high levels of rent extraction, as evidenced by the very high profitability of the banks. For the economy at large, this is quite costly and holds back on economic growth.
If the government wants to maximize the revenue from the privatization of the banks, it has to do so by enabling them to extract high future rent.
That however will be quite costly for households and firms, who would be much better served by a more diverse and competitive environment.
Because the government is now the effective owner of the banks, it has a unique opportunity to restructure the banking system to increase competition.
In turn that would increase future tax revenues from an economy that grows at higher rates
Maximizing the revenue from the privatization minimizes the future economic benefit from the financial system
Iceland is a very good example of the old saying that countries are always unhappy with their current currency policies. Even though it has used a number of regimes, it has been subject to continuous monetary instability ever since its got its own currency.
While there are no easy solutions, certain types of monetary policy are more likely to be stabilizing from others.
After the lifting of capital controls and when the subsequent imbalances have reduced, Iceland will have to establish a long-term monetary policy regime.
This subsection is copied out of my book Global Financial Systems.
Having a monetary policy objective, perhaps a formal inflation target, does leave open the question of how the central bank should meet the objective. One approach is the Taylor rule, proposed by Taylor 1993 whereby the central bank sets the nominal interest rates based on changes in inflation, output and possibly other economic variables. Under the rule, the central bank should increase nominal interest rates by more than 1% in response to a 1% increase in inflation. By having a formal rule, a central bank may avoid inefficiencies induced by a discretionary policy. Mathematically, we can state the Taylor rule as:
where is the target short–term nominal interest rate, the inflation rate (the GDP deflator), the desired rate of inflation, is the equilibrium real interest rate, an estimate of the logarithm of real GDP and is the logarithm of potential output, obtained by a linear trend. is the output gap. The parameters are restricted to be positive, and Taylor proposed setting them at 0.5.
In general terms, the Taylor rule seeks to apply negative feedback to the economy, increasing rates when either capacity is stretched or inflation is above target and reducing them when the opposite applies. This clearly matches central banks’ objectives in qualitative terms, but an important practical problem is the dependence of the rule on quantities that can only be approximated. In particular, GDP is only known with a considerable lag, and is subject to frequent revisions. Even in the long run, GDP is only an approximate measure of the economy.
Inflation targeting with the Taylor equation is not appropriate for Iceland.
Many central banks, explicitly or otherwise, use a form of the Taylor rule to set interest rates. It is however most suited for very large currency areas, such as the US, because it disregards the impact of interest rates on exchange rates. For much smaller countries, higher interest rates may lead to inflows of hot money and carry trading.
I discussed this in a seminar at the University of Iceland on June 6, 2008, titled “Lausafjárkreppan og hagstjórnarmistök”. The slides can be downloaded here (in English). I was quite heavily criticized by some central bank staff present at the talk for arguing that raising interest rates prior to the crisis were expansionary.
The evidence is increasingly strong. The Taylor equation works for a large economy, like the US, because when interest rates increase, the impact is confined to the domestic economy, increasing the price of money and reducing the money supply.
In a smaller country like Iceland, when interest rates are increased, the interest differential attracts carry traders, so hot money flows in, leading to a currency appreciation. This has several impacts:
The resulting vicious feedback loop is why inflation targeting using the Taylor rule is inappropriate for a country like Iceland.
Interest rates in Iceland have been very high for quite some time. Before the crisis, it was because the Central Bank used the Taylor rule to implement inflation targeting, and post crisis because the the IMF insisted.
If however inflation targeting with the Taylor rule is ill-advisable, then the conclusion must be:
Interest rates in Iceland are too high.
Inflation is costly, but so is fighting inflation.
While in larger and more diversified economies the cost of inflation outweighs the cost of fighting it, to me it looks like it’s the other way around in Iceland.
Therefore, Iceland would be better served accepting higher inflation if that ultimately is stabilizing.
A particular challenge for executing monetary policy is what to do with hot money inflows. Before the 2008 crisis, inflows frustrated policymaking, not the least because of the central bank’s use of the Taylor equation.
Inflow capital controls should not be implemented.
Going forward, policymakers may want to employ capital controls, but I think that is not advisable.
Suppose they require short-term flows to be deposited with the central bank at zero interest rates for a year. It would be very easy for any speculative trader to make a deal with a local exporter, so a speculative trade would be masked as an export related transaction. There are many other ways to achieve the same purpose and it can be very difficult for the authorities to prevent it. Just think of the Malaysian experience in 1997 and Mexico in 1992 with its Tesobonos.
The central bank will have to maintain a vast bureaucracy, monitoring currency transactions and punishing misbehavior. That would be expensive, error-prone and conducive to corruption.
Directly target the profitability and risk of carry trades.
By maintaining low interest rates, the profitability drops. By doing non-sterilized interventions, the risk increases. That might not be enough, but will go a long way and has the virtue of being self reinforcing.
There are several ways the authorities can implement monetary policy:
The monetary policy needs to target something, in recent years and decades it has been inflation, and before that exchange rates. Neither policy was successful. This suggests that a less orthodox approach to targeting might work better. In particular, policies should be countercyclical. So, a more flexible policy role would be to target stability.
This means not to use automatic mechanisms like the Taylor rule, and be willing to accept high inflation.
The authorities should not use capital controls to discourage inflows of hot money.
It is not advisable to use bank capital as a monetary policy tool. That means the same tool becomes triple use — macro prudential regulations, micro potential regulations and monetary policy. There will be constant conflicts us to which policy objective should prevail.
Make use of reserve requirements to control the money supply, give the banks of flexibility to fulfill the requirements in domestic or foreign currency. The interest on the reserve requirements should be zero.
Actively intervene in the foreign exchange market when money is flowing in. This of course is very costly as experience of Switzerland shows us, but some of those costs can be mitigated.
Do not sterilize the interventions. The Central Bank would find it difficult anyways given that it does not own much in terms of government bonds. By not sterilizing, the Bank keeps the costs down. Of course this is inflationary, since it increases the money supply. However, that inflation works as a cost to the carry traders, further dissuading them. Therefore, the need for such interventions will be much lower than it otherwise would be. Furthermore, the central bank can increase reserve requirements if it becomes wants to increase counter cyclicality.