---

"Being a Keynesian means being a Keynesian in both the good and bad times."

Andres Velasco (Finance Minister in Chile) [1]

It has been a while since I last had a thorough look at Chile (here and here); more specifically, the last time I had Chile under the loop was in October 2008 and thus around the time when the global economy was about to enter two quarters (Q4-08 and Q1-09) of absolute horror. Whether we are past the worst at this point in time is debatable and I am, personally, skeptical with regards the narrative of second derivatives and green shoots, but it is hard to deny that it does represent a narrative and a fairly strong one too. In this context I thought it would be interesting to have a look at Chile, how it has faired and how we can expect it to fair in the immediate future.

It will immediately become clear as we move forward through the data that Chile is a bit unique both in a global and most definitely so in a Latin American context. In this sense, and if not for any other reason, the following should confirm that although the global economy is in the midst of the worst crisis since the 1930s, there are some economies who are better positioned than others. In order to pin down some fix points from which to begin this analysis, it is interesting to go back to Q4-2008 and the note by Morgan Stanley analyst Luis Arcantales who pointed out that as the global economy was about to slide, it was Chile's time to shine. This analysis was echoed in the Economist's small article about Chile in which it is argued that Chile is cashing in the fruits of rigour.

The question is then; is this true?The analysis which follows supports this positive view on Chile and I thought it would be fair, at the offset, to identify the two underlying mechanisms for this position.

First of all, Chile has been saving for a rainy day and especially in the context of the copper windfall enjoyed in the past years, Chile have been acting with utmost prudence. Coupled with a big pool of sovereign assets/wealth tucked away in main state investment vehicles (SWF) this provides Chile with an enviable and essentially remarkably positive fiscal profile going into the crisis. The most important aspect of this strategy of prudence has been the joint commitment across political leaderships to maintain a structural fiscal surplus of 0.5% of GDP in order avoid the copper windfall from pushing Chile into a variant of the Dutch disease as well as of course as to lock in savings for rainy day. Between 1996 and 2006, Chile’s public balance averaged 1.5% of GDP and coupled with a substantial amount of the copper windfall parked in the SWF Economic & Social Stabilization Fund (FEES) it has granted Chile with a net debt position of -11% (i.e. a net credit position of 11%).

In addition to the story about the timely management of the Copper windfall, I have also emphasised the demographics of Chileand in particular the fact that the key working age brackets are still growing as a percentage of total population. In many ways, Chile is now moving on the outskirts of the so-called demographic dividend with the age group 25-64 still growing as a percentage of total population whereas the age group 25-44 is declining. It is an empirical fact that such favorable demographic momentum has a strong effect on macroeconomic performance; see e.g. Bloom et. al 2007 and Bloom and Williamson 1998.

However, with fertility coming in at replacement levels in these very years Chile now stands on the boundaries of the much debated second demographic transition (SDT) and it will be interesting to see just how Chile enters this second leg of the demographic transition (if at all). It is important to point out that the SDT is far from an inevitable process, but in it the light of the regularity with which life expectancy has continued to increased at the same time as fertility has steadily moved below replacement levels in one country after another, it is difficult to imagine that Chile won't also enter a new stage in its demographic transition. However, and whatever happens in Chile as we move forward it does not change the fact that Chile has the demographic winds blowing firmly in the back at the moment even if the direction is changing. The key will naturally be the extent to which Chile manages what comes next in terms of demographic evolutions.

Touched, but not Harmed?

Even with this set of formidable fundamentals the global economic crisis has not left Chile untouched. On a quarterly basis the third quarter of 2008 marks the last quarter in which Chile grew at the rates its citizens and policy makers have been used to over the course of the years of abundance leading up to the crisis. Since Q3-2005 the average growth rate of Chile's output measured by GDP was a remarkable 4.5% q-o-q, a figure which clocked in at a puny 0.2% in Q4-2008 and then on to a full blown contraction of 2.1% q-o-q in Q1-2009. In fact on an annual basis, Chile has observed negative growth rates since Q3-2003.

The central bank expects GDP for 2009 to hover around the 0% mark with -0.75% as a low point and the 0.25% as the corresponding best case scenario. This relatively bleak figure is produced by the expectations that domestic demand will contract at a rate of 4.7% of which the expected decline in gross capital formation of of -14.3% which contrasts with a 19.5% expansion in 2008.

This headline forecast naturally calls for all kinds questions not least the impending question, as it is being asked around the world, about the extent to which Chile will ever recover to observe growth rates it did before the global crisis. Personally, I believe that most analysts would agree that the script on 2009 as a horrible year is already written and the question now becomes; will 2010 be the year of recovery or will it be the year of disappointment as the boost from 2009's stimulus packages wane and it becomes clear that any kind of second leg with respect to a sustained pickup in global growth will be very tepid. I tend to lean towards the latter account, but it is also clear that the extent to which the global economy is able to limp forward, it will be economies such as Chile who will be doing a lot of the heavy lifting.

This particular view motivates a lot of what follows.

A Closer Look at Trends in Output and Activity

One way in which to differentiate the GDP measures fielded above is to have a look at GDP divided onto sectors to see how ouput in Chile has evolved over a wide array of activities as well as to compare this to some form of base value. As can be seen from the graphs to the right; I have chosen to focus the attention on cobber, manufacturing, construction, housing property, and financial services.

In the graph to the right the base value 100 is equal to the mean value of output over 4 quarters in 2003 measured at constant 2003 prices. For an economist with an inclination to base his analysis on underlying demographic parameters one thing immediately stands out. Indices for construction and housing property have hardly budged. This is interesting since the main driving force across of the real economic collapse across the globe is, in the case of many other economies, precisely driven by a collapse in these sectors. Now, whether this is because Chile did not entertain the same kind of bubble-like environment as elsewhere or whether it represents the fact that Chile's demographic profile would exactly lead us to the point that these precise sectors should be well supported by the underlying fundamentals I will remain silent. Clearly, it will be a bit of both, but it is a point worth remembering when talking about construction booms and bubbles; there is always an underlying capacity story underneath. This discussion is readily available in an Indian version concerning the risk of overheating which was debated furiously a while back.

Meanwhile, the general trend indicates that although there has been a notable drop in the constant price value (in mill pesos, 2003 prices) activity has not collapsed in any sense of the word and remain well above its base value. Now, there has of course been a decline and the jury is still out with respect to the extent that the decline will continue, stabilise or turn into growth. Most likely growth will resume its due course over the course of h02-2009, but as in all other places in the world it is the level of this growth which may ultimately surprise on the downside. One area where activity has markedly declined since the middle of 2008 is in the context of manufacturing and in this sense it is worth while having a closer look at the underlying pattern here.

If we start by looking at the manufacturing indices in the first difference (change) and represented through a 6-month moving average to try to smooth out the trend for the naked eye we observe the negative trend as it has grapped hold in the latter parts of 2008 and into 2009. However, we also observe that this does not like the horrible charts that we have seen e.g. in the context of the US, Europe and Japan. The average monthly rate of change in the general index through the 12 months ending April 2009 was -0.2% which is not exactly cataclysmic; in terms of the subcomponent the production of durables on the other hand decline at an average rate of a full 2% (mom) whereas the average change in the value of capital goods was 1%.

During the time measured in the graph to the right the general index peaked in March 2008 at 139.7 and bottomed in February at 112.8 after which it has recovered to 123.1 at the end of April. As noted, a large part of the drop in the latter part of 2008 and into 2009 was a sharp decline in the value of production of durables which fell (on an index basis) to a low of 65.9 in February 09. At this point in time the production of durables furthermore remain depressed relative its long term trend. Conversely, the value of production of consumer goods and capital goods have pretty much shadowed the trend in the general index; or more aptly, it is the relative stability of these two indices which have helped the general index to skirt what has been a sharp decline in the production of durables.

Finally and perhaps to end where I should have started it is worthwhile to have a look at the main index for economic activity in Chile (the IMACEC).

Looking at this index it is difficult not to conclude that Chile appears to have managed the initial stages of the economic crisis quite well. Surely, the index is down as one would expect but at this point at least, it does not appear to be a decline which will buck the general trend. The index peaked in June 2008 and has since fallen back 5% at the end of April. The most recent data however confirm that the slowdown is lingering as we approached the second half of 2009 with industrial production dropping 10.5% yoy in May prompting comments from central bank president Jose De Gregorio to note that nominal interest rates could be lowered further from its already low level at 0.75%. Moreover, the monthly GDP indicator showed that Chile continued to contract as we entered Q2 posting yoy 4.6% decline and with monthly inflation rates beginning to post negative readings policy makers and analysts close to Chile remain alert. As we have just rapped up Q2 in real time it appears that Chile is poised to surprise somewhat on the downside in terms of prior expectations.

The External Sector

The analysis of Chile's external balance and the country's currency is of course closely tied to the evolution of international copper prices as Chile is, by far, the world's biggest producer and exporter of copper.

Although copper prices have fallen back somewhat in the midst of the global recession relative to the average values through 2006-2008 they are still higher than they were at the turn of the century when the price of copper were below 1 USD/lb. In fact, the graph should make any trader look more than once since with the recent increase the price of Copper is very close to breaching the its 12 month moving average price although of course the strength of the global momentum in general will decide whether commodities, and thus Copper, will fly again. As an aside, it would be very interesting to run an analysis on the extent to which the recent move upwards in Copper prices has anything to do with the reports that China is stocking up on commodities(it does of course, but how much?)

The positive effect from copper on Chile's external balance has, at times, been coined as the copper bonanza and Chile's ability to manage this bonanza in a prudent manner is one of the reasons that the country stand out in the current environment. In general however, the composition of Chile's external balance look very much like one would expect of course that the current account has been in surplus since 2004 due to the positive impact from the trade balance and thus net exports of copper. Thus, up until the advent of the financial crisis Chile's current account was characterised by a positive trade balance which outweighed a negative income balance to produce a consistent current account surplus. This changed in the latter part of 2008 where Chile posted a current account deficit in Q3 and Q4 as copper prices plummeted and exports in general fell. Basically, the trade balance withered away into a small deficit and with a continuing negative income balance, Chile found itself in need of external financing for the first time in 5 years. It also pushed the current account deficit into deficit for the full year 08 and the central bank, rather surprisingly, expects 2009 to see another CA deficit. I say surprisingly here since Q1-09 has so far posted an overall CA surplus worth 639 billion USD driven by a strong trade balance (mainly due to a plunge in imports and higher Copper prices). In any case, it is difficult to imagine that Chile will any problem financing a current account deficit of the magnitude the central bank is forecasting at 1.8% of GDP in 2009.

Turning the analysis to the currency it is interesting to observe that last time I looked at Chile inflation was running close to 10% and with nominal interest rates below the inflation rate the Chilean economy was experiencing negative real interest rates. In the context of the currency this meant that just as we were rounding up Q3 2008 the Chilean central bank decided to hold back on its frequent endeavors into the market to stem the rate of appreciation of the Peso against the USD. Endeavors, which by the way, have been unable to buck the overall trend in appreciation of the CLP ever since 2003 against the USD.

Of course, events had it in Q4 2008 that markets were to experience a significant amount of stress and rising volatility which sent the Peso down against the G3 currencies where it is only now recovering. In the context of the stress encountered in the market and seeing that the spread on Chile's sovereign debt increased less than the average in Latin America (and Asia) I argued that perhaps this was a sign that Chile's currency would not be hit as hard, in the context of increasing volatility, as its emerging market peers. My argument in a nutshell was that since the Peso was amongst one of the best performing emerging market currencies against the USD (back in April) this was perhaps due to the relatively high standing Chile had with international investors. Stefan Karlsson would have none of this however arguing in stead that the relative strength in the context of Chile's Peso was to be found in relation to the increase in the price of Copper. I conceded that Stefan was right in so far as goes the obvious fact that Copper is a very important driving force for the Chilean Peso regardless of whether investors were also targeting Chile as a relative safe haven amongst emerging markets.

However, in the spirit of good argument I decided to let me and Stefan's arguments suffer the, not always flattering, test of empirical validity. To that end I cooked up the following small model;

Y = a + b1X1+b2X2

Where Y is the exchange between the Peso and the USD (quoted directly), X1 is the price of Copper, and X2 is the sovereign spread. I use monthly data from Jan-00 to May-09 for a total of 112 observations and as per convention I am estimating this model in the first difference to avoid issues of stationarity [2]. Given the hypothesis one would expect a negative sign for X1 (i.e. an increase in the price of Copper is associated with an appreciation of the Peso) and a positive sign for X2 (i.e. an increase in sovereign spread is associated with a depreciation of the Peso). The estimation (with OLS) returns the following result;

Y = 0.0016 - 0.16X1 + 0.09X2 + ut [F = 33.25, R-sq = 0.38]

Now, both variables (X1 and X2) are significant at 1% [3] and thus I am inclined to stick my neck out a little bit more vis à vis Mr. Karlsson and conclude that the extent to which investors see Chile as a relative safe haven amongst emerging markets will in turn make Chile's sovereign debt spread increase less relative to its peers in relation to market turmoil which, in turn, has a measurable effect on the exchange rate.

Don't worry, this will be the first and last regression analysis you see in this note and just to sum up; Copper does matter for Chile and with net revenue expected to drop 69 percent this year to $1 billion from $3.2 billion in 2008, it will have a noticeable impact on Chile's economic performance although I need to emphasise that, to my mind, Chile posseses sound fundamentals which move far beyond the benevolence of its Copper ressources.

Employment

In terms of the labour market Chile cannot escape the fact that the crisis has taken its toll. The latest figure for April has the unemployment rate running at 9.6% which makes it almost certain that it is above 10% in the time of writing. 10% hardly constitute a dramatic number in a relative context (although of course it is big in an absolute sense), but given the fact that Chile entered the crisis running at 7-8% the lagged effect of the recession on the labour market may push the unemployment rate to uncomfortable levels which is sure to become a big topic for the elections later this year.

The number of persons employed peaked in August 2008 at 6.693.400 persons and has since declined to 6.574.500 persons for a total loss of employment of 118.900 people in April 2009. At the same time the registered number of persons in the labour force increased by 120.140 people from 7.196.110 to 7.316.250. These figures highlight one of the challenge with having a large and growing labour force in the sense that you need to maintain momentum in order to be able offer the jobs which the people rightfully demand.

Of course, a growing labour force is a good thing in itself, but in the current environment we should not rule out the case that it can become a source of "unrest" and fierce political debate. Should the employment situation continue to deteriorate on the margin (that is unemployment reaching some 15%) it will be very interesting to see how this drives the discourse in the upcoming elections.

Policy and Inflation

As noted, the last time I had Chile under the loop the central bank perceived the risks to economic stability in a wholly different light than it does now. At the time, inflation was running at some 10% on an annual basis and the central bank was busy moving up nominal interest rates. That has changed now.

Chile's central bank is formally targeting an inflation rate of 3% and just as it was running way above this target in the period leading up to the crisis, so has it plummeted accordingly and is currently running at negative values on a monthly basis. This has prompted the central bank to lower rates to an unprecedented level of 0.75% in June and most analysts expect another nudge downward come the July session (the graph to the right plots the interbank rate). If this turns out to be the case, the central bank will have lowered interest rates by 7.75 % over the course of the last 6 meetings. Just as it has been the case with other more prominent central banks, the Chilean derivative is trying to steer expectations in an environment where long term yields have begun to inch upwards to reflect the solidification of the second derivative discourse. In general, the central bank is tracking inflation closely with its target interest rate as can been in the graph to the right.

On the fiscal front Chile is in a much better position than most. Alongside the measures taken on the monetary front the government has, so far, initiated US $4 billion package of government spending and tax cuts. According to the budget office the budget deficit will amount to 4.1% of GDP this year, a position one finds it difficult to believe that Chile will have trouble financing. On June the 15th Chile's fiscal authorities announced a bond issuance worth $ 1.7 bn as well as its intent to use $4 bnfrom its offshore savings to fund spending.

Not too Shappy

All in all this does not look too bad now does it? In many ways I agree with CitiGroup's research department as they wrote in their latest overview of the Latin American economies;

We believe that the Chilean economy is one of the best positioned to capitalizefrom a global recovery. The openness of the Chilean economy made it one ofthe most vulnerable to the global slowdown, certainly after Mexico. But thestrength of its domestic fundamentals helped the economy withstand the globalshock.

Clearly, there are downside risks here and these come mainly from any adverse shocks Chile might suffer from another global fallout or simply the risk that global growth won't recover to the extent many are currently expecting. Yet, it is important to point out here that Chile's relative strength has two sides. On the one hand there is no doubt that the presence of Copper and the important of this commodity in the global value chain as well as the sound management of the windfall from this is very significant. On the other hand I have also, as per usual, emphasised demographics as a key variable and specifically that Chile is still riding the waves of the demographic dividend, or more aptly the afterburner of this process. In fact, what is important for Chile at this point is to lock in the favorable path by avoiding that fertility falls too much below replacement level. If Chile succeds in this, it may truly turn out to be an example to follow on more than one front. 

---

[1] - I distinctly remember that he has been quoted for something like this, but I don't remember the exact wording. 

[2] - I use the following formula ln(t0/t-1).

[3] - If you run regressions as single linear models in turn with X1 and X2 respective as explanatory variables this pattern is repeated with almost identical R-sq values albeit somewhat higher for Copper prices.