The sharp collapse in the rupee in late 2011 created a sea of red ink on corporate balance sheets. Despite this trauma – and similar ones suffered with gruesome regularity over the years – companies continue to access external commercial borrowing (ECB) aggressively. ECB taken in 2011-12 (April-January) was nearly $30 billion (Rs 1.52 lakh crore), up over 18 per cent on the $26 billion (Rs 1.3 lakh crore) in all of 2010-11.
In most cases, the major part – over 60 per cent – of ECB remains unhedged as companies try to time the market to capture significant savings in borrowing costs. Some of these efforts end up in tears, and all of them go through balance sheet trauma from time to time.
So, why do companies persist with this approach?
In some cases, they believe a foreign exchange (FX) liability is effectively a natural hedge for exports. This fallacy is reinforced by both the central bank and commercial banks, which are, of course, simply looking to ensure that FX outflows are covered by inflows so there is no cash risk.
In reality, a medium-term FX liability only provides a natural hedge to a medium-term FX asset. Exports are, by their nature, short-term assets and so are not meaningfully hedged from a business standpoint by an FX loan. One side of the equation invariably ends up subsidising the other (either the loan will subsidise exports, if the rupee strengthens, or vice versa) and it becomes difficult to understand where the business margin comes from. (Having said that, the value added in the company’s business is, indeed, a medium-term FX asset, which can be naturally hedged with an FX liability.)
In most cases, of course, companies don’t hedge because they find the cost of hedging much too high. If you borrow three-year dollars today at, say, Rs 50 to the dollar, hedging it out completely at a premium of six per cent a year is equivalent to buying dollars in 2015 at 60 to the dollar. While it is certainly possible that the rupee could hit 60 some time in the future, the natural inclination is to believe that you will be able to exit at a better price before the loan becomes due.
In actual practice, things often don’t work out that way.
We ran a study comparing the 100 per cent-risk (unhedged) and zero-risk (fully hedged) cost of three-year amortisation dollar-denominated loans, each loan drawn down at quarterly intervals from January 1, 2006. We found that the average FX cost of the loan if it were left fully open (1.47 per cent per annum) was, indeed, lower than if it had been hedged fully on day one (1.85 per cent).
However, the averages hid a wide variation. For the unhedged loan, the worst cost was 8.81 per cent, while the best cost was -2.86 per cent; for the hedged loan, the variation was much lower with the worst cost at 3.63 per cent while the best cost was 1.03 per cent.
Clearly, the worst cost of the unhedged loan, after adding in about 4.5 per cent (the cost of fully hedged Libor+250 basis points), worked out to more than 13 per cent — not particularly attractive compared to most companies’ rupee borrowing costs. On the other hand, the best case was excellent, providing an all-in cost of rupee funds of less than two per cent — perhaps a worthwhile risk-to-reward ratio.
Of course, staying fully unhedged is hardly an approach anyone would recommend — and certainly not company’s boards and other stakeholders whose business it is to worry about risk.
Given this, and the fact that few companies are intuitively comfortable with hedging fully, we developed a structured approach to liability management, using as a base our proprietary model for short-term FX risk, which has been delivering well-defined value to dozens of companies for several years now.
Over the test period, the approach has performed extremely well — at 0.81 per cent on average, the model improved on the cost of being either fully open or fully hedged by 0.66 per cent to 1.04 per cent per annum, respectively. The range was still quite wide — the worst cost came in at 7.65 per cent, more than a percentage point better than the worst unhedged cost, but quite a bit (four per cent) worse than the worst fully hedged cost. However, the best performance – at 2.96 per cent – was excellent, a little better than the best unhedged loan but substantially better than the best hedged loan. Of course, tweaking the hedge signals could provide different ranges, depending on how much risk the company was willing to carry.
In the best incarnation, this approach could provide the best of both worlds — a full one per cent per annum better on average than being fully hedged, and a much better risk/reward than fully open.