Saturday, October 15, 2011

Reining in the inflationary dragon

A lot is being written about inflation in India today. I thought it's worth writing about the fascinating insights into inflation that come from focusing on the distinction between tradeables and non-tradeables.

What is a tradeable

A tradeable is a product which can be transported across the world at relatively low cost. As an example, steel is tradeable while cement or paint are mostly non-tradeable barring special short-hop opportunities like Gujarat-Karachi or Amritsar-Lahore or Calcutta-Chittagong or Trivandrum-Colombo.

Steel is a nice tradeable that one can think clearly about. There are no barriers to the movement of steel worldwide. Hence, there is only a world price of steel. The quoting convention used worldwide is to express the price of steel in USD. The price of steel in India is thus the world price of steel multiplied by the INR/USD exchange rate, plus a markup for freight (The cif/fob ratio).

If there is a customs duty of (say) 10%, then the price of steel in India is 1.1 times the world price of steel expressed in rupees. For the rest, nothing changes when a customs duty is introduced. Gram for gram, every fluctuation in the INR/USD or the world price of steel shows up in the domestic price of steel.

Non-tradeables are things like cement (which are hard to transport) or haircuts (which are impossible to transport).


Before we can analyse and control inflation, we must measure it well. Inflation is defined as the rise in the price of the average household consumption basket. The CPI is the best measure of inflation in India.

Everything in the CPI basket can be classified into the two categories: tradeable vs. non-tradeable. As a thumb rule, WPI non-food non-fuel is a rough measure of tradeables inflation. Fluctuations in food and services prices, which make the CPI diverge away from WPI non-food non-fuel, are a measure of non-tradeables.

Year-on-year inflation reflects an averaging over 12 months. If you want to get a faster sense of what is going on, you need to look at point-on-point seasonally adjusted changes. These yield early warnings of inflation, which are 5.5 months ahead on average. Such data is updated every Monday by us. The shift from y-o-y inflation, to p-o-p SA inflation, is a free lunch in measurement and monitoring.

The WPI is a useful database of many price time-series in India. But the overall WPI is useless in thinking about inflation in India: there is no household in India which consumes the WPI basket.

The use of WPI inflation, and the exclusive use of y-o-y inflation, are litmus tests of professional competence in the Indian landscape.

The function of the central bank

The job of RBI is to deliver low and stable inflation: to deliver y-o-y CPI inflation of between 4 to 5 per cent.

They have failed in this task. From February 2006 onwards, in every single month, y-o-y CPI inflation has exceeded 5 per cent. This is an important time for introspection at RBI and outside it. What have we done wrong, in the structuring of RBI, which has got us into this mess?

It is useful to think of this as a principal-agent problem. The people of India are the principal. RBI is the agent. The principal hires the agent and gives him resources. In return, the agent has to be held accountable. Delivering low and stable inflation is the accountability mechanism. It is a quantitative monitorable measure of the performance of the central bank. That we have sustained failure on this function, from February 2006 onwards, suggests that we should be modifying the nature of the contract between the principal (the people of India) and the agent (RBI).

How RBI can influence the price of tradeables

RBI has absolutely no say on the world price of steel. In that sense, the prices of tradeables are beyond the control of RBI.

When RBI raises the interest rate, more capital comes into India, which tends to give an INR appreciation, thus making tradeables cheaper. Thus, an RBI rate hike does impact upon the domestic price of tradeables.

It is also worth pointing out that the central banks of most major countries are high quality inflation targeters. They deliver on their mandate of delivering low and stable inflation. As a consequence, inflation in the global tradeables basket tends to be low and stable. Tradeables prices are a helpful source of price stability, most of the time.

(That a large part of the CPI basket is tradeable, and seemingly beyond the control of the central bank, is no excuse. There are dozens of high quality central banks visible in the world, with very large shares of the CPI basket in tradeables, who are delivering on inflation targets. We in India should not accept excuses).

How RBI can influence the price of non-tradeables

Non-tradeables reflect aggregate demand and aggregate supply in India. RBI can influence these by raising or lowering the short-term interest rate. When interest rates are made slightly higher, household consumption and investment demand are slightly lowered.

A critical feature of non-tradeables inflation is expectations. If people expect 10% inflation, they tend to wire high price rises into their negotiation of wage and other contracts. This generates inflationary momentum. Particularly in a place like India, where the institutional structure of monetary policy is primitive, economic agents have little confidence in the ability of policy makers to rein in inflation. As a consequence, inflation is highly persistent. Once high inflation sets in, economic agents expect high inflation to continue. There is a great deal of momentum in inflation.

For years now, some economists have argued that inflation will subside by itself. It will not. Inflation does not mean-revert to the target zone of 4 to 5 per cent by itself. We are now in a trap of high inflationary expectations. This structure of expectations will need to be broken. This can happen in two ways. RBI needs to turn a new coat, and convince people that it now cares about inflation without any other conflicts of interest. And, rate hikes have to take place.

There are two paths to inflation control: changing the structure of expectations and reducing aggregate demand. The former is almost a free lunch. It only requires institutional change. The latter is hard work; it inflicts pain.

What about supply factors?

Some argue that supply bottlenecks in India - such as hideous rules about mandis - are the cause of inflation.

The trouble with this explanation is that the supply bottlenecks have always existed. They have existed in high inflation times and in low inflation times. It is, thus, not possible to claim that supply bottlenecks have caused the inflation crisis which began in February 2006.

Can rate hikes deliver inflation control?

When C. Rangarajan was RBI governor, there was an inflation crisis, and rate hikes did deliver on inflation control. The phase of price stability ushered in then lasted all the way till February 2006. This shows us that even in India, it can be done.

We have to remember that in his time, the monetary policy transmission was much weaker than what we see today. With a bigger wall of capital controls, domestic rate hikes did not deliver inflation control by impacting on the INR (through higher capital inflows). With a smaller and weaker Bond-Currency-Derivatives Nexus, the monetary policy transmission from the short rate into aggregate demand was inferior, then. Yet, he got it done.

Conversely, with a very primitive financial system and monetary policy transmission, the central bank of Zimbabwe delivered a nice hyperinflation. We can quibble about the potency of the monetary policy transmission, but we should not doubt the ultimate domination of monetary policy in shaping inflation. In the long run, little else matters in shaping inflation.

Part of the story of the 1990s lies in clarity of purpose at RBI and policy credibility. Rangarajan's period had good quality speeches, which did not dilute the message on inflation control as the dharma of the central bank. In contrast, in recent times, RBI has repeatedly written low quality speeches. To an expert reader, they have conveyed the lack of knowledge on monetary economics at RBI. To the non-expert reader, they have waffled on the subject of taking responsibility, and have encouraged the average economic agent to think that high inflation is here to stay.


  1. Dear Ajay,

    Pardon my limited understanding of the subject, but your last para seems to suggest that RBI can talk the inflation down.

    Surely, apart from monetary policy, fiscal policy initiatives are very important too. Even if it means unpopular measures and more pain.

    Naresh Tejwani

  2. I am a consumer of any noise coming out of RBI as a demand from my profession. But the business media (both MSM & social) seem to have missed the biggest slip in recent times: substantial devolvement of G-Secs in the auction concluded on 14-Oct-2011: INR 4000 Cr of devolvement in two of the most active securities (~45% of the issue size).

    Here we have a situation where the Government is refusing to disinvest its PSU holdings on the back of poor valuation, unable to meet its revenue targets, exceeding its expense budget, overshooting the borrowing plan and saying "not any rate can be accepted as cut-off". RBI has been preaching about the market determined rates (wrt INR depreciation) but where is the same diktat when it puts a upper cap on yields (or so it tries) at which Government borrows. I cannot fathom a situation where the borrower whose debt-equity is rising by the day, wants to borrow more but refuses to pay more for the incremental debt. For reining in inflation and anchoring expectations, the demand side needs to be taken care of. With Government being the biggest (and the most reckless) spender, if the borrowing rate does not go up simultaneously, in my understanding, the seriousness and veracity of RBI's handling of inflation control seems thoroughly out of place.

  3. i dont get the expectations theory at these guys at the RBI have so much power outside of the printing press they wield?.
    really? just by writing speeches which ajay shah would approve,inflation would go down?. i mean,forget RBI,why dont influential people on TV have a better chance to reign in inflation then? just get rakhi sawant or shahrukh khan to reduce inflation.which businessmen or entrepreneur pays attention to RBI speeches? which housewife?
    dont the huge increase in monetary base in the western world have any bearing on us? especially given that rupee is not convertible? we are importing inflation .that is the west's gift to the rest of us.

  4. It's not just by writing good speeches. It's by behaviour which abides by the good speeches.

    The motto of a good central bank is: "Say what you will do, and do what you just said'.

    The scarce currency is that of credibility. Over the years, RBI needs to build up credibility. At present, with five years of failure on inflation, this is sorely lacking.

  5. I'm not convinced, especially with the argument on non-tradeables. Yes, interest rate can be used to influence consumption to an extent, but that is why there is a measure called elasticity of demand. Food/vegetable price inflation has been severe since 2006 (the time frame over RBI policy is being criticized) and demand for basic foods is pretty inelastic. So I don't think the supply-side constraints argument is closed. Ideally, I'd like to see a plot of real per capita income growth Vs time(for say the last decade), and I'd like to see that further split into real per capita income growth across the lower, middle, and upper income brackets(since lower class income growth would spur a greater demand for basic foods). Then one should show agricultural output Vs time. Then we can conclude based on the relationships in those data whether there is or isn't supply side constraints.

    As for tradeables, see this for more:
    The above may not tell the whole story. But its a factor to consider. Arbitrarily raising interest rates will absolutely kill investment without necessarily hitting at the root cause for the inflation.

    I will post this as a comment may be. I'm sure he knows more than I do, so I'd like to see how he addresses these questions.

  6. Dear Ajay,

    Why the focus on consumer prices? What about asset prices which have a direct impact on the economy. With losses in the stock and bond markets and an induced recession the impact is clearly observable in the data currently rolling out.
    Is it not the responsibility of the RBI to address both issues?
    Other tools are required and a focus on interest rates is not sufficient to address the issues.
    What say you?


Please note: Comments are moderated; I will delete comments that misbehave. The rules are as follows. Only civilised conversation is permitted on this blog. Criticising me is perfectly okay; uncivilised language is not. I delete any comment which is spam, has personal attacks against anyone, or uses foul language.