Unlike most commodities that are provided by private actors competing with each other, a currency is provided by a monopolist. The only institution that is allowed to produce laris is the National Bank of Georgia (NBG).
The task of the NBG-monopolist is made difficult by various peculiarities that cannot be found in other markets. First of all, lari once injected into the economy are not consumed or used up. They remain in the economy “until the cows come home”, as they say in Scotland. This is starkly different to what we observe in most other markets. If bakeries interrupt their production of bread, after a short period of time the bread that was circulating around is eaten.
Then the ratio of supply and demand in the market decreases and the bread price increases. In contrast, if the National Bank of Georgia stops printing money, the amount of money circulating around remains the same. Hence, it is very costly for the NBG to reduce the supply of money, as this does not mean that the “production of lari” has to be interrupted but lari have to be bought back, depleting scarce foreign currency reserves.
On the other hand, an ordinary monopolist typically faces production costs, and in this respect the NBG has an easier life. It is almost costless to increase the supply of money by starting the printing press.
SHOULD THE NBG STEP IN?
Many Georgians were unsettled by the sharp depreciation of the lari vis-à-vis the dollar that occurred within the last months. Indeed, as the graph shows, since the beginning of November 2014 the value of the lari, measured in dollar, moves constantly downwards. On the first of November one could buy one lari for approximately 57 dollar cent in international currency markets, while today one has to pay only about 47 dollar cent (about a week ago the situation was even worse, when a lari was traded for just 44 dollar cent). In relative terms, moving from 57 cents to 47 cents is a devaluation of almost 20%.
There was a lively discussion about whether the NBG should use its foreign currency reserve “ammunition” for buying back lari and thus reducing the supply. Such an intervention was demanded by Georgian politicians, but history is full of examples which show that it almost never works. Faced with international exchange rate dynamics, central banks, even if much bigger than the NBG, are largely helpless. It is a war that cannot be won, and the foreign currency reserves get lost underway.
In January 2014, the Russian central bank held more than 500 billion dollars in foreign currency, and to this day they used approximately 120 billion dollars to buy back rubles. Yet did it help? From January 2014 to today the ruble lost approximately 50% of its value, much more than the lari in the same time. Even the Russian central bank, really having a lot of foreign currency in store, is impotent in view of the massive loss of trust in the ruble throughout last year.
When there are fundamental downward forces, interventions on part of the central bank can usually delay or slow down the depreciation for some time, but is it worth the money? A famous case is the Mexican peso crash in 1995. When the peso came under pressure, many US pension funds were engaged in Mexico, and they feared a loss in the value of their (financial) investments made across the border. Political pressure was exerted on the Mexican central bank to delay the peso crash, and Mexico was provided with a multi-billion dollar bailout package. These measures, however, die not prevent the overall depreciation of the peso – it was just delayed, giving the US investors the necessary time to disengage. The episode was not to the advantage of Mexico, as later they had to pay back the bailout package given to them, while their currency had crashed nonetheless (a detailed account of this and similar cases can be found in Lester C. Thurow’s The Future of Capitalism, Nicholas Brealy 1997).
Even more problematic would it be to peg the lari to the dollar in a fixed exchange rate regime. This would invite currency traders to apply their standard speculation scheme. Worldwide, they could coordinatedly borrow lari in huge amounts and then sell these borrowed laris for dollars at the fixed exchange rate. The central bank guarantees the exchange rate, so it has to sell dollars for lari. Yet in the process, the foreign currency reserves of the central bank melt away, and at some point the central bank would run out of reserves and could not buy further lari. At that point, the lari crashes, allowing the speculators to buy back the laris they owe to their creditors for just a fraction of the dollar price they had received when selling the laris previously. If the central bank really defends to the bitter end but is finally defeated, all its currency reserves are pocketed by the speculators. Even the heavy-weight Bank of England fell victim to this scheme, when in 1992 they had pegged the British pound to a basket of other European currencies, mainly the German mark. Despite trying to defend their exchange rate, they were finally defeated, and in the process transfered huge amounts of money to speculators, among them George Soros who made about one billion dollars in this attack against the pound.
To sum up, there is not much what one can do on the supply side of lari. If one wants to support the exchange rate, one has to stimulate demand. Yet as it turns out, this is also not easy.
WHY DO PEOPLE (NOT) DEMAND LARI?
The three classical functions of money are medium of exchange, store of value, and unit of account (see, for example, Burda and Wyplosz’ Macroeconomics: A European Text, Oxford University Press 2005). While the last function does not drive demand for money, the first two are relevant for how much lari (or any other currency) people want to hold. If one plans expenditures in lari, be it for investments in Georgia or just buying Georgian products, one needs to have lari.
Some people have argued that there is less demand for lari because the investment climate in Georgia has deteriorated. This is countered by those who say that foreign investments did in the past not account for a large share of the demand for lari, and hence investor confidence could not play a big role in the lari depreciation.
However, it is difficult to say how the price of the lari reacts to a change in demand. Perhaps, having a little more investor confidence and therefore a little more demand for lari would make a considerable difference in the price. Moreover, the demand for laris is not governed by past activities but by future plans and expectations about Georgia’s economic wellbeing. If people demand lari as a medium of exchange, they plan to do transactions in the future – the medium of exchange purpose of money is all about future transactions. Statistics of the past do not capture these expectations.
If one wants to make the lari attractive as a medium of exchange, there is not so much one can do, yet a sound economic policy which creates a good investment climate would be advantageous. Some policy measures of the last two years, like the new immigration law, were not helpful in this respect. In this context, it is important to keep in mind that an economy which heavily depends on foreign investment is much more sensitive to policy mistakes. In Germany, they recently introduced a minimum wage which came with ridiculous bureaucratic requirements (employers have to write down the exact time, minute by minute, at which each employee works), and there is a plethora of other economically questionable decisions that are made in Germany in every year. Yet Germany’s economy is diversified and strong, and it takes many mistakes to bring the economy down (not to say that this will not happen at some point). Georgia, however, mainly depends on foreign capital and is therefore much more fragile. The decision not to invest in Georgia is a rather easy one, given that there are many comparable countries competing for investor money.
Finally, the medium of exchange purpose of the lari is related to what people expect they can buy with these lari, i.e. what and how much the Georgian economy will produce in future. Growth prospects depend on reasonable economic policies, and one may wonder whether the last two years were used optimally to foster growth in Georgia.
However, the biggest driver of the lari depreciation is the fact that people do not perceive lari anymore as a good way to store their value. This is of course interconnected with the medium of exchange purpose – if there will be many people demanding lari for future transactions, lari will be a reasonable way to store one’s value.
There is a global movement towards storing one’s value in dollar, illustrated by the fact that not only the lari depreciated but also many other currencies of developing economies. From October 2014, the Armenian dram devaluated by around 16%, the Turkish lira by around 11%, and the Azerbaijani New Manat, which until recently was defended bravely by the Azerbaijani central bank, just in the last few days lost more than 20% of its value. The differences in these values may be due to different investor confidence and future expectations, yet by and large it is a global trend beyond the control of a small country like Georgia.
This shift in preferences to not store so much value in lari anymore is like a natural disaster. One can hope that the crises in Ukraine and Russia as well as other geopolitical threats will lose intensity, but as long as Georgia’s economy is in a fragile state, which it will be for decades to come, it will be exposed to the capers of the global economic climate. The only safe harbor would be to join the Euro area.
Comments
"...lari have to be bought back, depleting scarce foreign currency reserves." – there surely are more ways to decrease the money supply, which do not involve using foreign currency reserves at all.
"However, the biggest driver of the lari depreciation is the fact that people do not perceive lari anymore as a good way to store their value" - that would be true, had not the dollarization of the economy (both loans and deposits) been decreasing steadily over the past decade. Slight increase in dollarization, reported in January, is explained by depreciation (no accounting for the change in the exchange rate has been done in any sources quoting said increase) and by government removing certain deposits of LPPLs to the Treasury (net 200 mln GEL).
The biggest problem is the imbalance in inflows and outflows of foreign currency, manifested in decrease in exports and remittances, stagnation in tourists' numbers (slight decrease vs January last year) and increased or non-decrease of exports (let's see what February says). Hence increased demand for USD and depreciation.
There are not really other ways to get money out of the system. If you increase the interest rate, more money will be deposited at the central bank and it is not circulating around, but the money is still in the system. It is just temporarily taken out but still belongs to private actors who can at any point in time decide to bring it back into the economy, for example by exchanging it for dollar.
Regarding your second point, I am not talking (only) about dollarization, i.e. Georgians storing their value in dollar. Also foreigners, who may have hold some of their value in lari (due to future plans and expectations) are now exchanging them for dollars.
What you identify as the "biggest problem" is certainly correct, but what you mention are just the symptoms. My article searches for the underlying reasons for those symptoms, i.e. explaining why the outflows exceed the inflows.
My reply should have got here, sorry
And why would dollarization cover only Georgians? Of course it applies to anyone holding deposits in resident banks.
I wasn't speaking only about Georgians and foreign residents to Georgia. Anybody who had plans with Georgia may have stored lari. One may hold lari even for purely speculative purposes. Lots of Forex traders around who are speculative-minded.
Here's what you said: Regarding your second point, I am not talking (only) about dollarization, i.e. Georgians storing their value in dollar. Also foreigners, who may have hold some of their value in lari (due to future plans and expectations) are now exchanging them for dollars.
This certainly means that you didn't consider foreigners, also exchanging lari for dollars, as part of dollarization problem, does it not?
And you cannot have any reliable data on forex traders holding lari (apparently outside Georgia, otherwise it is still part of dollarization), let alone exchanging that lari into dollars, so they simply are out of the picture. My argument's still valid.
No, I wasn't talking about dollarization. Never had that in mind even. You brought that up. It is an issue which also affects the demand for lari, but I did not think about dollarization as an important driving force behind the reduced demand for lari.
Why should it be relevant whether or not there is reliable data on forex traders?
Dollarization is the only reliable indicator showing shift in trust from lari to dollar. Everything else is just a speculation.
If there's no reliable data - what's the point of speculating on forex traders? None whatsoever.
Not every valid economic argument must be empirically testable. I am making deductive arguments, the standard way to argue in economic theory. I make assumptions, some of them implicit, and derive my conclusions from there.
One can make assumptions based on existence of Russell's teapot. That wouldn't make these assumptions valid.
True, of course you could question the validity of the assumptions I make. But just saying there is no data available does not invalidate my argument.
1. Of course there are - you sell securities, t-bills, CDs, logic being the same as in the case of FX. And, btw, is money not deposited at the central bank after selling foreign currency? There's no difference at all.
2. There's no run from lari at the moment - that's for sure. I'm not really sure where you got the idea from.
3. Of course it's not a symptom. How can an a drop of remittances be a symptom of loss of trust in lari - it just can't.
Also, what you are saying has nothing to do with outflows or inflows. It might have to do with the change in portfolio composition of agents within the country, true. But it doesn't explain (and how could it?) why people send less money or buy less Georgian products.
1) If the NBG sells securities it already owns, like T-Bills, it is doing the same as buying lari with their cash. Whether the NBG holds their currency reserves in cash or in T-Bills does not make a difference.
If you mean that the NBG would sell its own bonds, this would be similar to increasing the interest rate. The money is withdrawn temporarily, but when the debt matures, it is paid back to the private actor. Anyway, is there any central bank in the world which is issuing bonds? Never heard of anything like that and would be surprised if the NBG would be legally allowed to issue their own bonds.
2) You know, the price of a currency is -- like anything else -- governed by supply and demand. If the supply remains the same, as it is the case, then what MUST be responsible for the price decline is reduction of demand. This means that there is of course a "run from the lari", if you want to express it that way. There is simply no other possibility.
3) The loss of remittances belongs partly under my second category, "medium of exchange". The recipients of the remittances would have got dollar amounts and then exchanged them for lari to make purchases. This demand for lari has now gone down due to a lack of remittances.
1. Of course it does. The same logic could be applied to fx reserves - just a temporary move away from lari currency.
Who said that the debt should ever be repaid? You never need to repay the debt, so long as you pay the interest. And NBG has certificates of deposit, btw. Not a bond but a monetary instrument of similar effect.
2. The price of lari is not an exchange rate - it's an interest rate. Exchange rate is a relative price of lari and is governed by supply and demand of both lari and dollar. As simple as that.
3. Simply not true. The recipients of remittances still have the same demand for lari, they would have, had their remittances stayed the same - these things are not related at all. They are just able to sell their remittances at a higher prices - to those who need those remittances and whose demand for remittances hasn't decreased yet (importers etc.)
1) No, you can't apply the same logic to foreign currency reserves. If the NBG buys lari for dollars, these lari are permanently out of the economy.
Except for English Consols, I do not know of any debt that is never repaid. If the debt has a maturity, however, there is a moment where the creditor can decide not to renew the debt but to, say, buy dollars instead. So, the money is only temporarily out of the system.
Indeed, CDs are just a way how the central bank can borrow money from private actors, so what I said about bonds applies to CDs as well.
2) For the price of everything you need a numeraire, i.e. a unit of account. This numeraire can be chosen freely, you can for example take dollars (which would be the natural choice if you are speaking about the price of a currency). Of course, the interest rate is not the price of the lari. The interest rate is the price of risk taking and deferred consumption. If the interest rate goes up, then, everything else being equal, one is willing to pay a higher price for my willingness to take risk and to wait with my consumption.
3) Okay, you can call the decline of remittances the cause of the reduced demand for lari, and I would be fine with that. But it translates into people having less demand for lari as a "unit of exchange", so it is not something not covered by the analysis I offer.
One interesting thing I noticed is that in 2013 Georgia re-exported automobiles almost worth the same as it imported, whereas in 2014 the gap between import and re-export increased by $190 million, probably due to the car import ban in Azerbaijan. So if trade deficit increased by $600 million from 2013 to 2014 the share of car re-export decrease is pretty big in it.
Did this affect Lari depreciation?
Should Georgia have done the same as Azerbaijan?
Hi gg12,
You are absolutely right. This explains why in the last months of 2014 the trade balance became even more negative. Clearly, this put additional pressure on the lari. In my opinion, however, these country-specific reasons are not significant in view of the general preference to hold value in dollars (which affected most currencies around the world).
What do you mean by "should Georgia have done the same as Azerbaijan"?
Regards,
Florian
1. No, NBG can use these lari to buy dollars again. Your logic is kinda one-sided.
Technically, you don't need to repay the debt, you can always roll it over with another creditor. So no, not necessarily temporarily.
2. So if we take Somalia Shilling as a numeraire, its recent appreciation would mean that the demand for dollar has fallen and people start to move to shillings? Or, if we take a period where lari was depreciating towards dollar, but appreciating towards euro - that means reduced confidence in euro? Please. And interest rate is often called price of money, I'm not inventing things. Risk taking etc - are part of value of said currency.
3. Erm, no. Demand for lari does NOT decrease with drop in remittances. Nor should it. So you analysis still not relevant.
1) To later increase the amount of lari later is a free decision of NBG. If they don't decide in that way, the money won't flow back into the economy.
Assume the NBG wants to roll over your debt but there is nobdoy willing to hold lari debt. Whenever debt matures, the creditors are taking the money and exchanging it for dollar. There is no guarantee that you can roll over the debt. In any case, you may have to pay such high interest rates that through those interest rates the amount of lari circulating increasing.
2) You are right about that. The lari may be a very goo performer if we take Somali shilling as our numeraire.
3) I think my argument is waterproof. Let's agree do disagree.
1. Aren't both actions free decisions of the NBG? What is the difference?
I said, if creditors are willing. If they are not, that would indicate much larger problems for the economy than we have now. Way, way larger.
2. So there's no validity in your argument, since the numeraire is arbitrary. But if it's not (and it's not - we need a valid numeraire to be able to draw valid conclusions) - that supply of dollar is as important for the exchange rate, as that of lari. Even more important, btw.
3. Look. I am a Georgian person living solely on remittances. I also live on subsistence level and my remittances are just enough for that. So my demand for lari is 160 lari per month (or whatever the minimal basket is). How the fact that people send me 50 dollars instead of 100 is going to decrease my demand in lari?