A year after RBI cut its policy repo rate steeply by 1.25 percentage points (to 6.75%), the (new) 10-year benchmark Government Security (GSec) is currently trading at 7.82% (the old one is 8.0%), more or less at the same levels at the start of the current rate-cut cycle in January 2015. State bond auctions saw yields for 10-year paper reach as high as 8.88%. The spread of the 10-year GSec in normal times is usually 40-50 basis points over the repo rate. The benchmark paper is important, helping to fix rates in varying magnitudes across the spectrum of interest rates. This large (and increasing) divergence is one of the causes behind the stickiness in repo rate transmission to bank lending rates.
This divergence has been attributed to many causes. Bank demand for GSecs, as part of Statutory Liquidity Ratio (SLR) holdings, has not kept pace with the supply of paper. This is partly due to the low growth in bank deposits, which tracks bank advances. Regulatory changes have also contributed: A phased reduction in the SLR ratio to 20.25% over the period till March 31, 2017, as well as corresponding cuts in the Held to Maturity (HTM) ratio (which enables banks to offset the provisioning for market risk). This is reducing the risk appetite for investment in GSecs.
In FY16, foreign portfolio investors (FPIs) increased their holdings of government securities very marginally compared to the previous fiscal year. There is reportedly a preference for state government bonds (called State Development Loans, or SDLs) among another set of significant buyers of government debt—insurance companies and provident funds—who look to boost their interest yields, since weighted average of SDLs’ yield rose from about 0.3 percentage points over their central government counterparts in mid-2015 to over 0.7 in January 2016.
This last phenomenon—the increased issuance of SDLs over the past 3 years—had sneaked past rather quietly, till the time this began to impact the GSec yield curve recently, partially due to FPIs’ lack of interest.
Gross issue of SDLs have increased from Rs 1.6 lakh crore in FY12 to an expected Rs 3.1 lakh crore in FY16, with the sharpest annual delta of Rs 70,000 crore in FY16. While part of this, naturally, is due to higher redemptions, these have been nowhere close to the fresh issuance. As a proportion of central government paper, net SDL issues have increased from about 30% in FY12 to an expected 62% in FY16.
Why have state borrowings increased so much? This is a complex set of issues, not amenable to generalisations across states, but the following is a broad narrative of some of the reasons.
Analysts report that the proximate reason is an increase in interest payments, but this is not the key problem. Interest payments on market loans were up from Rs 62,000 crore in FY12 to a budgeted Rs 92,000 crore in FY15.
The share of total interest payments to total revenues has actually come down from 13% to 11.2%. Interest on market loans have remained stable at 5%.
As the accompanying chart shows, there was a sharp jump in states’ fiscal deficit, from 1.9% of GDP in FY13 to 2.5% in FY14. This was largely due to a fall in the revenue surplus of states.
Part of the larger fiscal deficits was funded by a drawdown of the accumulated cash surplus with states beginning FY14, but the rise necessitated an increase in borrowings. With the cash having been drawn down in FY14, the same level of fiscal deficit in FY15 consequently required a further increase in market borrowings, from Rs 1.8 lakh crore in FY14 to Rs 2.5 lakh crore.
The presumption is that this phenomenon has since accelerated, in FY16. The initial presumption was this might have been caused by lower than budgeted transfers of central funds devolution to states, but the finance ministry has clarified that the budgeted devolution of funds is on track till February, and will be completed in March 2016.
The following assessment is mostly conjecture, since many details are not available. Based on a sample of 12 states whose FY16 Budget Estimates are available, one reason for the larger deficits was a sharp drop in non-tax revenues, a large part of which is “grants from the Centre”. These grants are usually matching grants, contingent upon allocations by states. Till about FY14, and maybe even till FY15, drawing down on cash surpluses, states were able to make these allocations. Once the cash surpluses were drawn down, the ability to draw on central grants fell sharply.
States seem to habitually overestimate central grants from Central Budget Estimates, and actual grants are even lower.
One proxy for substantiating this is a look at states’ holdings of 14-day T-bills, which had risen by Rs 28,000 crore in FY13, then shrunk by Rs 43,000 crore in FY14, up by Rs 10,000 crore in FY15 and again shrunk by Rs 34,000 crore in FY16 (till February 12, 2016). The stock of investments in 14-day T-bills have halved from a high of R1.42 lakh crore in FY13 to Rs 75,000 mid-Feb. One reason for these fluctuations might be a change in allocation of small savings collections.
The higher borrowings are uneven across states. Of the Rs 70,000-crore jump in net borrowings in FY16, just short of Rs 50,000 crore was by seven states, Rs 35,000 crore by four. The reasons are unclear. There is a partial overlap of the states with the largest increase in borrowings with the ones with the largest losses of their respective electricity discoms, but the correlation remains unexplained (higher subsidies might be an explanation). As a proportion of funding their fiscal deficits, the market borrowings are also widely dispersed, with FY15 borrowings of the largest borrowers ranging from 50% to 115% of their FY15 fiscal deficits.
With contributions from Abhaysingh Chavan
The author is senior vice-president and chief economist, Axis Bank. Views are personal