Cash holdings from land sales, larger than warranted by economics, might be adding to demand
The companion piece to this article, a few days ago, sought some insights into why India?s inflation, initially manifest largely in food inflation, but now gradually disseminating to other components in the consumer price index, remained so persistently high, particularly in an environment where growth (for whatever reason) has slowed so dramatically and swiftly. This is an issue that policymakers will have to grapple with, to diagnose the causes and set in place mitigating measures. This is particularly important for monetary policy, tasked with price stability as a primus inter pares objective. Coincidentally, even as this column is written, one of the global credit ratings agencies, generally perceived to have taken a more benign view of India?s macroeconomic fundamentals, has also reportedly pointed to high food prices as a negative for sovereign credit ratings. The implication for a monetary policy stance is explicit and immediate.
We had earlier looked at aspects like the factors usually attributed to keeping demand, especially rural demand, high: government subventions, rising minimum support prices, rising rural wages, inadequate improvements in agri productivity, etc. Each of these has undoubtedly played a role in pushing up aggregate demand, but still leaves a sense of collectively being an insufficient explanator for the magnitude of rise in prices.
One driver we had touched upon was the effect of land procurement?increased cash generated from acquisition of rural land, for roads, SEZs, industrial projects, etc?in boosting disposable incomes of landowners. It is simply impossible to put a hard number to the quantum of these deals in the absence of primary surveys. The effects of these transactions would differ markedly by geography, proximity of projects, anticipated development, and so on.
Yet, a broad aggregate number can be estimated based on certain assumptions on real estate transaction multipliers. These transactions, even if done through cash, do involve the transfer of cash. And we have a pretty accurate estimate of cash in the system, since this is printed by RBI.
One indirect means of assessing this quantum of cash used in real estate (and, of late, for gold) is through the increase in currency with the public, adjusted for estimated cash required for conducting normal economic activity. Total cash in the system is currently a bit more than R11 trillion (lakh crore), of which 95% is cash with public. As a share of GDP (see chart 1), these balances have increased from about 10% of GDP in early 2000s to 11.8% in FY09 before dropping slightly to 11.4% in FY12. Over the years 2003 to end-March 2012, these cash holdings have increased by about R7.5 trillion.
Is this level of cash holdings warranted by the increase in economic activity over the years 2003 to 2012? Since 2005, the extent of transactions conducted through electronic channels have risen from about half of all (electronic and physical) transactions to over 90%. Adjusting for this increase, the amount of cash actually required to service economic activity ought to have been around 6.9% of GDP in FY13, rather than the reported 11.4%.
This translates into about R4 trillion more cash in the system than would be warranted by increase in economic activity since 2003 (which is about the time the real estate boom started). Again, we emphasise that these are ballpark numbers, and much more detailed work is required to finesse the estimate based on segregation of bulk electronic versus retail payments.
But the excess R4 trillion might be a working figure to begin with. Is this being used for real estate transactions? Bank lending for real estate (housing plus commercial real estate) rose from an outstanding total of less than R50,000 crore at March 2003 to over R5,00,000 crore in March 2012. Assume housing finance companies (HFCs) would have lent another 75% of this amount. Total formal finance, then, would have been upwards of R8 trillion. Normal loan to value (LTV) for formal housing finance is about 80%, i.e. a home buyer would need to put in 20% of own equity, with another 80% as loan. Factoring in the extent of unaccounted transactions, this LTV might drop to as low as 30-40% of the value of real estate transactions. Assuming a 40% LTV, the increase in imputed value of real estate transactions would have been over R20 trillion.
This prima facie seems to have an implicit multiplier far larger than feasible, given the estimated excess cash of R4 trillion (even if we attribute part of this to higher gold holdings). Even for speculative transactions, this would imply a transaction velocity much higher than that suggested by reality. Therefore, either the magnitude of excess cash is much higher (less probable) or the de facto LTV is higher than assumed above.
Within the band of error of the above back-of-the-envelope calculations, there will be associated implications for the various monetary indicators, which are inputs into monetary policy making. For instance, the drop in financial savings, with consequences for bank deposits (and manifest in the drop in M3 growth rates) has increased the cost of funds for banks, thereby impeding transmission of monetary policy easing signals.
The author is senior vice-president, business & economic research, Axis Bank. Views are personal