Clemens Grafe of UBS -

Now that the financial sphere has normalised, with the ruble getting mildly stronger against the currency basket, interest rates having fallen sharply to the support level and inflation having dipped, the next question is how quickly Russia's real economy will react to this normalisation. Clearly, this will depend to some extent on the oil price, the outlook for which remains uncertain.

Still, it is hard these days to find anybody who argues that it will fall below $43 per barrel Brent (the level at which the Central Bank of Russia would potentially change the ruble band) for sustained periods. Most market participants seem to see it remaining around current levels for the next 18 months, with the risks skewed, if anything, to the upside.

Looking at GDP growth, which was reported at -9.5% on year in the first quarter after 1.2% positive growth in the fourth quarter, looking for signs of any "green shoots of recovery" in Russia seems more like wishful thinking. However, we think the 9.5% contraction does not accurately reflect the underlying sustainable growth trend. As we have written previously, domestic demand (excluding inventories) has been contributing around -4.7% to the negative GDP growth. With imports down in volume by 22%, we think net exports will ultimately contribute positively to GDP growth after a -0.7% contribution in the first quarter, which, in our view, was driven by Gazprom's abnormally dismal export performance of down 56% in volume terms. Fully 4.1 percentage points of the 9.5% fall in the first quarter was thus explained by inventories, which should largely be a one-off. In fact, even the negative contribution is at least partially due to inventory adjustments. While gas in storage tanks in Western Europe fell by 11bn cubic metres (cm) in the first quarter of 2008, it fell by a full 28bn cm in a year later, explaining probably half the 28bcm fall in Russian gas production in the period.

With gas prices now finally falling and oil prices (to which gas prices are linked with a 6-9 month lag), we think gas inventories are likely to be rebuilt. We therefore think that the underlying decline in GDP was more of the order of 4-5%, which is in line with our -4% forecast for 2009. Irrespective, it is clearly the case that the economy continued to contract in the first quarter, and there was no sign of a stabilisation on a quarterly GDP basis.

Data a guide to past, not future

In the mature economies, analysts search for turning points by looking at the quarter or year before growth rates of seasonally adjusted data, which is seen as giving the best assessment of current economic conditions. Turning points and recessions are also timed on this kind of data, not on year-on-year data. Thus, when our US colleagues say the US will show positive growth in the third qurater, they mean seasonally adjusted economic activity in the US will be higher in then than in the second, although in year-n-year terms the economy is still expected to be down in the third.

Unfortunately, Rosstat does not publish seasonally adjusted data, which makes comparison with other countries less straightforward. It is also not that easy to accurately gauge the equivalent dynamics from year-on-year data either. For instance, a constant year-on-year growth rate is in most cases not clearly an indication of a stabilisation of economic activity. In the four quarters to the second last year, the Russian economy expanded by 8%, or around 2% a quarter. Thus, a constant year-on-year growth rate between the first and second quarters would indicate an expansion of the economy of approximately 2% on quarter, rather than just stabilisation.

Financial markets are forward-looking, so what they are most interested in is the current growth rate of GDP as the best indication of the expansion (contraction) of future activity.

We have seasonally adjusted the main Russian macro series ourselves, using the openly available algorithm of the US Census Bureau, to get a better idea of where we are in the cycle. As the charts above indicate, GDP declined by a seasonally adjusted 9.5% on quarter in the first quarter after declining 1.25% in the fourth; the Russian economy has hence shrunk by around 10.7% since the peak.

Now, this clearly all refers to past, yet what we are really interested in is the economy's current run rate and whether there are any signs of stabilisation/ growth. Below we have plotted the seasonally adjusted growth rates for industrial production, retail sales and other indicators to get some idea of where the Russian economy might go from here.

Both manufacturing and mining grew in seasonally adjusted terms in the first quarter. However, in the case of manufacturing, this growth occurred from a level 18% lower than prior to the downturn. Given that oil production is actually up on the year, the impact on the mining side is hardly visible in terms of the level.

In the manufacturing sector, the picture clearly differs substantially across sub-sectors. Basic metals is now producing at a level 30% below prior to the crisis, while food production is already virtually back to where it was prior to the downturn (in seasonally adjusted terms food production rose by 15% annualised in the first quarter, essentially reversing the fall in the previous three months).

Seasonally adjusted production levels in March were between 5% (food) and 50% (machinery) below the average level in the year to October 2008 (see chart below). Thus, the sectors that have seen the sharpest contractions are those that are either geared towards investment (machinery), consumer durables (transport) or export (metallurgical, chemical and timber).

Even the sectors geared to investment are arguably mostly reliant on exporting industries. Thus, without an upturn in exports, it is difficult to see an upturn in manufacturing, given that these sectors are the heavyweights in manufacturing.

Fortunately, however, the mining sector - with the exception of gas - is not contracting, and should thus ultimately lead to a manufacturing recovery.

Domestic industries such as food instead seem to be working at capacity utilisation rates that are close to those prior to the crisis. While this means they most probably cannot expand very quickly, given capacity constraints, their cost advantage following the ruble's depreciation should incentivise expansion.

Unlike on the production side, there are no signs yet that consumption is stabilising. Even in seasonally adjusted terms growth is clearly still negative. In terms of money growth though, the worst seems to be behind us. M2 in seasonally adjusted terms is rising again, in our view a reflection of money returning into rubles.

Constraints on lending

We are often asked whether it is not possible to link lending growth to investment growth. This sounds far more straightforward than it is. According to official statistics, companies' earnings have been by far the most important source of capex finance, followed by budget funds; only around 10-15% of capex in large enterprises is funded by loans and capital markets. In other words, profitability is a far greater source of capex than lending. This is why we expect margins and profits to be the main driver of any recovery, rather than bank lending.

Banks are not lending at present, and have effectively frozen their corporate loan books. Instead, they are concentrating on building up buffers to absorb expected loan losses, and this is clearly restricting the economy. However, we believe that bank finance - when it is available under normal circumstances - is mainly used for working capital requirements, not capex. Thus, when bank funding freezes up, it has a very immediate impact on the economy.

That said, we believe corporates will have used the last few months to replenish their cash reserves and are meeting an increasing proportion of their working capital requirements internally. Also, while we think the flight of $160bn of Russian capital has been the main factor behind the country's economic downturn, the fact that so much Russian money fled before also means that there are clearly funds that can return to Russia once confidence returns.

Given the credit squeeze, the Russian market for assets has clearly become more rather than less active; banks are taking over equity of debtors, while corporates with debt they cannot refinance are selling assets to companies with cash, and so on. How long this process will continue depends to some extent on whether assets have bottomed out or not. Once asset prices stabilise, the market tends to clear more easily, as it becomes easier for banks to evaluate borrowers' creditworthiness, while buyers have no incentive to wait before coming to the market etc.

Unfortunately, there are not many asset prices for which market data is available. Clearly the stock market has turned, and is up 70% year-to-date, but it is less clear how quickly this will feed through to other assets. The only other series we have are real estate prices. Moscow real estate prices are available on a weekly basis from IRN (www.irn.ru). As the chart below shows, they are still falling by any metric (USD/sqm, Rub/sqm and CPI-deflated Rub/sqm). However, in dollar terms at least, the rate of decline is already slowing down very significantly. So, how long might it take for the housing market to stabilise? Alternatively, how much more downside is there to house prices?

We consider it unlikely that house prices will fall significantly below the price/income ratio seen in 2004, prior to the lending boom and prior to the commodity boom. Oil prices (even deflated) are still far higher than in 2004, and the economy is very unlikely to shrink back to that level. Thus we would think that there is at most another 10-15% downside to house prices in ruble terms, following the 23% fall in January to April. At the current rate of decline, this will maybe take another 2-3 months.

As the chart below shows, the price adjustment in Russia has been far more abrupt than in Western countries. In the US for example, house prices have been falling since July 2006, and in local currency terms are down by 32%, a decline that Russia will, in our view, have achieved in just six months. While this means more short-term pain, clearly the period until things stabilise will also be significantly shorter.

What type of recovery?

In the developed world, the general view is that the downturn will be stopped by the fiscal and monetary response, and consumers starting to spend again. In China, hopes of a recovery rest on credit policies and infrastructure spending. We believe that in Russia the dynamics is likely to be quite different. For consumption to lead the economy out of the woods would require a level of private-sector wealth, acting as a temporary source of demand, that is simply not there. The wealth of the general population is just not sufficient to act as a buffer against economic fluctuations.

While the fiscal balance has clearly shifted sharply from a surplus of 7-8% of GDP in the first half of 2008 to a deficit of 3.3% of GDP in January-April 2009, this is mainly due to the reduction in oil revenues (7% of GDP), while the government's 4.4% of GDP crisis package in 2009 is mainly geared towards a stabilisation of the banking sector and dampening the social impact of the downturn. Infrastructure spending has, if anything, been cut.

This makes us firmly believe that the recovery in Russia will be driven by the supply side, and it is production indicators that therefore need to be watched in order to discern the turning point. In short, we believe that corporates will cut costs and raise margins to levels where it becomes attractive to expand again. In this respect we find it highly encouraging that manufacturing profits have recovered markedly in the last two months.

We have been repeatedly asked if it isn't possible to construct a leading indicator for Russia that would give clients an indication that we are over the worst. Even after looking at a lot of data, we have to admit that we cannot really find any series that would beat the oil price as an early indicator. Clearly, we are not alone in this. Even the OECD leading indicator does not really beat oil in terms of gauging the turning points of the cycle.

The leading indicators (which include the oil price) really turned in August 2008, as did the oil price, while industrial production really turned in June 2008. Thus, the OECD indicators would not have been very useful in helping us to predict the downturn. In our view this is not because of the OECD methodology, or their choice of indicators, but just because the gyrations in the oil price were firstly extreme and secondly pretty difficult to predict, as Russian equity market performance demonstrated as well (the slight lead of the market versus the oil price was probably more due to the conflict in the Caucasus than a genuine turn in market expectations).

While clearly as analysts we do not like to say this, we have to admit that we feel there is little we can add to the identification of turning points that would not be easily observable in the oil price. On the other hand, this also tells us that we should not overcomplicate things, and just accept the oil price for what it generally tells us, ie. that the Russian economy is turning.

We have collected the indicators that we think are most relevant for tracking economic activity forward (other than retail sales and industrial production data) and have aggregated that data into an index.

Our composite indicator consists of the following eight series, which we aggregate with equal weighting after normalising both their means and their variances (the same methodology as used by the OECD): (1) the level of orders and stocks (inventories) in manufacturing as the OECD reports them; (2) the change in the CBR's FX reserves as a sign of capital flows; (3) m/m seasonally adjusted M2 growth as an indicator of domestic activity; (4) seasonally adjusted manufacturing profitability; (5) the 5-year CDS spread as an indicator of risk; (6) the real oil price in Rubles as an indicator of profitability in commodity sectors and for other discussed reasons; (7) the m/m change in real house prices in Moscow; and (8) the real three-month Mosprime rate.

As the charts show, these indicators have clearly turned (with the exception of the indicators on orders and stocks) and this is not only because oil prices have turned. Constituent indicators dynamics is presented in the charts below. All indicators are normalised to the mean of zero and the variance of one.

Send comments to The Editor