Envisioning the INR as a floating exchange rate

by Rounak Hande, Rajeswari Sengupta and Ajay Shah.

The Question

A central question in macroeconomic policy is the exchange rate regime. In the long-run, India's economic strategy should move to a combination of inflation targeting, floating exchange rate and an open capital account. In more than three decades since the economic reforms of 1991, only one of these milestones has been achieved--RBI today is an inflation targeting central bank. For IT to be fully effective, it must be accompanied by a floating exchange rate. However, we in India are used to the idea that the RBI actively intervenes in the FX market to stabilise the USD/INR rate. It is important to ask, What might a genuine floating exchange rate look like if the RBI did not intervene? In other words, If RBI were to the USD/INR what SEBI is to the Nifty, what would that world look like?

When it comes to government price controls on commodities, e.g. wheat, there is a ready way to visualize what a reformed India would look like: the Indian price of wheat would be the world price of wheat. But what about the exchange rate? If the required reforms took place, and we got to a market determined rupee, what would it be like? In this article, we present a reasonable depiction of what the exchange rate regime would be like, if the RBI did nothing on the currency market. This also helps us understand how much currency volatility Indian firms and households need to prepare for if the RBI were absent from the market.

The period of INR as a float

When we look back into India's history, we find that there was one period when trading by the RBI on the currency market dropped to near zero levels. We treat this as a natural experiment to gain insights into what the INR would look like without government control. Our first task is to establish the start and end dates of that period.

We start with three long time-series graphs: (i) RBI's spot market trading volume in USD, (ii) RBI's spot market trading volume relative to reserve money, and (iii) RBI's open position on the currency forward market.

Figure 1: The long time-series of spot price trading volume by RBI, in billion USD

Figure 2: The long time-series of spot price trading volume by RBI, expressed as per cent of M0.

Figure 3: The long time-series of the RBI's currency forward position, in billion USD.

In all these graphs, we can spot one remarkable period, from June 2009 to October 2011, where an important reform of the exchange rate regime took place, and the RBI stepped out of the currency market. Let's zoom into that period. To obtain greater clarity, we focus on the period from June 2006 to October 2014, adding three years to each side.

Figure 4: Spot price trading volume by RBI, in billion USD (June 2006 to Oct 2014).

Figure 5: Spot price trading volume by RBI, expressed as per cent of M0 (June 2006 to Oct 2014).

Figure 6: RBI's currency forward position, in billion USD (June 2006 to Oct 2014).

In these pictures, we see a middle period -- 28 months from June 2009 to October 2011 -- when currency trading by the RBI was very low. The RBI's trading volume in these months was not always 0. In choosing these endpoints, we set a limit where the RBI's gross monthly trading volume stayed below 1 percent of M0.

Examining the characteristics of this period gives us insights into what a floating exchange rate in India might look like.

Characteristics of the INR as a float

In this section we describe the characteristics of the INR in the period from June 2009 to October 2011. For the sake of comparison, we use the methodology described in Sengupta and Shah (2026) to establish the dates of two other exchange rate regimes. We will now focus on three such regimes:

• The natural experiment of the INR as a float: 1st June 2009 to 31st October 2011.
• The recent period of a tight USD peg : 1st September 2023 to 16th December 2024.
• The present exchange rate regime: 27th December 2024 to 28th August 2026 (latest available data).

For each of these periods, we examine (a) The volatility of the USD/INR rate (b) The parameter estimates obtained from the exchange rate regression (see Google colab notebook associated with Sengupta and Shah (2026)) and (c) Deviations from market efficiency as seen in variance ratios.

Metric	Float (June 2009–Oct 2011)	The USD peg (Sept 2023–Dec 2024)	Current ERR (Dec 2024–Aug 2026)
Volatility:
USD/INR vol (%)	7.40	1.43	4.98
The exchange rate regression:
USD	0.66***	0.89***	0.79***
EUR	0.20**	0.05	0.11
JPY	-0.15**	-0.01	-0.08
GBP	0.04	0.05	0.25
R-sq	0.75	0.98	0.76
RSE	0.77	0.18	0.67
Variance ratio tests:
VR(5), daily	0.99	0.67	0.96
p value	0.82	0.03	0.61
VR(4), weekly	1.07	0.64	0.76
p value	0.93	0.03	0.14

We summarise our findings as:

• In the popular discourse on the INR, a lot of attention is given to the raw USD/INR volatility. At present, it is running at 4.98 percent. During the period of the USD-peg, it had fallen to 1.43 percent. We see that under the float, it was 7.4 percent. In other words, if the RBI did not intervene in the currency markets, the USD/INR volatility that the economy could experience is around 7-7.5 percent. Later in the article, we speculate on how things might work out if the INR were to return to a float under the present conditions and we argue that the volatility could be lower.

  The numbers also suggest that during the current exchange rate regime (27th December 2024 to 28th August 2026), the machinery of the RBI's currency policy seems to have delivered only a small decline in volatility of about 2.4 percentage points on an annualised basis.

• In the exchange rate regression, during the period of the USD peg, the USD coefficient was statistically significant with a value of 0.89, and no other currency was significant. At present, the USD coefficient has come down to 0.79, but still, none of the other currencies have statistically significant coefficients.

  In contrast, during the period of INR float, the USD coefficient was smaller at 0.66, and other currencies were significant too. This suggests that the Indian economic engagement with the outside world is not merely with the US. In the float period, USD, EUR and JPY were all statistically significant, with coefficients of 0.66, 0.2 and -0.15. This gives us an undistorted sense of the currencies that matter for the Indian economy.

• Another important statistic from the exchange rate regression is the residual standard deviation (or RSE, residual standard error). It shows the size of the prediction error of the exchange rate regression. A lower RSE means the model fits the data well. Therefore, during the period of the USD peg, the residual standard deviation was only 0.18. In contrast, under the INR float the RSE was 0.77. In the current regime, the RSE stands at 0.67.

• In the exchange rate regression, during the period of the USD peg, the R-squared was 0.98. A high R-squared value implies that almost all of the variation in the INR was accounted for by the currencies in the regression model. At present, the R-squared has fallen to 0.76. During the INR float, the R-squared had a very similar value, 0.75. In other words, despite active trading by the RBI in the currency market in the present period, there is not much of a difference in the R-squared.

  This yields insights into deciphering a floating exchange rate regime from the data. A floating exchange rate does not necessarily mean an R-squared value close to 0. It means that the central bank does not intervene and lets the exchange rate respond freely to market forces. In a floating regime, the R-squared value can still be high because it reflects the natural, underlying co-movement of the rupee with major currencies of the world (and not just the USD) under conditions of globalisation. India is deeply interconnected with these countries through trade and financial flows, and is exposed to the same global shocks. Some co-movement is therefore entirely consistent with a genuine float.

• The variance ratio test is a simple tool to examine serial correlations. A floating exchange rate is expected to be an efficient market, with no discernible serial correlation, and no exploitable profit opportunities for trading based on time-series characteristics. This does work out correctly in the float period. In the daily data, the 5-period variance ratio was 0.99, and indistinguishable from 1, whereas in the weekly data, the 4-period variance ratio was 1.07 and indistinguishable from 1. Under the USD peg, the two variance ratios were 0.67 and 0.64, with statistically significant deviations from non-forecastability. In the present arrangement, the variance ratio at 4 weeks is away from 1.

A useful variant of the exchange rate regression, introduced in Kumar et. al. (2020), differentiates between the USD coefficient when faced with a USD appreciation vs. a depreciation thereby highlighting asymmetric intervention by the RBI. We now turn to these estimates.

Metric	Float (June 2009–Oct 2011)	The USD peg (Sept 2023–Dec 2024)	Current (Dec 2024–Aug 2026)
USD (App)	0.52***	0.87***	0.68***
USD (Dep)	0.83***	0.91***	0.88***
EUR	0.20***	0.05	0.12
JPY	-0.15***	-0.01	-0.09
GBP	0.04	0.05	0.24
R-sq	0.77	0.98	0.77
RSE	0.74	0.18	0.66

During the period of the USD peg, it is not surprising to see statistically significant values of the USD coefficient close to 1, for both USD appreciation and USD depreciation. This makes sense because when the RBI is pegging the INR to the USD, it is expected that currency interventions would take place on both sides of the market, regardless of which way the USD is moving. In the present exchange rate regime, the INR responds more strongly to a USD depreciation (with a coefficient of 0.88) than it does to a USD appreciation (with a coefficient of 0.68). This implies that the RBI now intervenes asymmetrically, letting the INR move more freely when the USD appreciates (i.e. the INR depreciates) but managing the INR more when the USD depreciates (i.e. the INR appreciates). This is consistent with existing studies documenting the RBI's asymmetric intervention patterns (Patnaik and Sengupta, 2022). RBI prefers buying dollars (preventing INR appreciation) over losing reserves (preventing INR depreciation).

Interestingly however, we find asymmetric coefficients in the floating period too, with a response of 0.83 when the USD depreciates but a coefficient of 0.52 when it appreciates. This is puzzling because in a float, there should be no asymmetry between these coefficients, both of which should be equally low. Further research is therefore required to understand the market-based sources of this asymmetry.

Conclusion

In the strategic view of macroeconomic policy, the long-run answer for India lies in graduating from one milestone -- inflation targeting -- to two more milestones -- a floating exchange rate and an open capital account.

At every stage in the journey of Indian economic reforms, the prospect of getting the government out of price determination has raised alarms in the minds of some people. When the proposals to remove price controls for steel or cement were made, there was shock and unhappiness in the minds of many people. These things are often easier done than said, because the price system works rather well. It solves the resource allocation problem, and prices move continuously in a way that provides good incentives to private persons.

In this article we have shown one tangible period, of 883 days, in which the RBI stayed away from the currency market and there was a genuine floating exchange rate. This period can be utilised for many other research projects. Using the insights from this period, we are now able to offer a thumb rule to judge the extent of government management of the exchange rate in India in terms of three numbers. For example, we can compare the values observed today of (i) the USD/INR volatility of 4.98 percent, (ii) the USD coefficient of 0.79, and (iii) the RSE of 0.67 vs. the values observed during the float: (i) the USD/INR volatility of 7.4 percent, (ii) the USD coefficient of 0.66, and (iii) the RSE of 0.77. This gives us a sense of how much government control of the exchange rate is present today.

This natural experiment, of a country that graduated to a floating exchange rate and then retreated from it, gives us insights on how to interpret the estimates from the exchange rate regression and the toolchain of Zeileis et. al (2010).

Looking into the future, when economic policy reforms take place in India, we believe the USD/INR volatility under a true floating exchange rate will be lower than this value of 7.4 percent, for two reasons:

1. There is one important difference between the float period of 2009-2011 and the future: Inflation Targeting. That RBI movement to a floating exchange rate was incomplete because it was not accompanied by inflation targeting. In some sense, that was a particularly unfortunate event as the rupee lost its nominal anchor during that period. In the future, things will be better because now the nominal anchor is 4 percent CPI inflation.
2. Another important difference concerns the liquidity of the USD/INR spot and derivatives markets. In the 2009-2011 period of INR float, these markets were less developed. We estimate that in that period, the total turnover(onshore and offshore) was about USD 40 billion per day. By now, things have improved, with a huge increase in INR activity outside India. Now the total turnover (onshore and offshore) is about USD 140 billion per day. This bigger market delivers greater stability. Hence, we can speculate that in the future, things will be better in terms of USD/INR volatility.

References

Kumar, S H, Balasubramaniam, V, Patnaik, I and Shah, A (2020), "Who cares about the Renminbi?", Working Paper, December 2020.

Patnaik, Ila and Rajeswari Sengupta (2022) "Analyzing India's Exchange Rate Regime", India Policy Forum, National Council of Applied Economic Research, vol. 18(1), pages 53-85.

Sengupta, R and Shah, A (2026), "Words and deeds in the Indian exchange rate", The Leap Blog, May 19, 2026.

Zeileis, A, Shah A, and Patnaik, I (2010) "Testing, monitoring, and dating structural changes in exchange rate regimes", Computational Statistics & Data Analysis, Volume 54, Issue 6.

Rounak Hande and Ajay Shah are researchers at XKDR Forum, Mumbai and Rajeswari Sengupta is a researcher at IGIDR, Mumbai.

Watching markets work: Structural change in the Nifty implied volatility

by Surbhi Bhatia, Anjali Sharma, Susan Thomas.

When options are actively traded, a new fact about the economy is revealed in option prices: what option traders think that the market volatility is going to be. This derived volatility forecast is called 'implied volatility' (IV). The Nifty IV measures the Nifty volatility that market participants forecast for the coming two to four weeks.

This was not visible without derivatives trading. Nifty options trading began in year 2000. This gives us close to 20 years of data on market volatility forecasts in 2018. We used this history to calculate the daily IV time series. We used the official daily settlement prices of eight at-the-money options, for both the call and put options.

In the early years, the options market was quite illiquid, so the IV every day could be quite noisy. On some days, the computed IV was wrong in overstating the views of market participants, and on other days, the computed IV was wrong in understating the true views of market participants. If we wish to know the IV at a point in time, we have to bring options market liquidity integrally into the calculation. For the present purpose, we find it useful to use monthly averages, which we feel are adequately reliable. Our estimates of IV are protected, first by the averaging that goes into the official settlement price, and second when we average the daily values across the month.

Such a time series allows us to examine the views of traders about future volatility through these years. Were the traders' expectations of market volatility the same today as when the markets started? Did the market volatility rise to reflect the turmoil in the markets during the global financial crisis? How do IV values compare across the two banking crises, of 2001/2002 and 2017/2018?


Our first step is to look for structural breaks. If there are different IV regimes in the series, this test will identify time points at which the regime shifted from one to another. We use the Perron-Bai algorithm as implemented by Achim Zeileis et. al. (the `strucchange' R package). This yields the following result:

This shows four phases of the story of Nifty IV:

June 2000 to April 2006: The first regime is from the outset till April 2006. This period started after the nuclear tests of 1998, when Nifty had gone to its lowest value (887). IV surged when the UPA won the elections and the stock market crashed on 17 May 2004. The average IV in this period was 21.1%.

April 2006 to August 2009: The second period ran for around three years, from April 2006 till August 2009. Roughly speaking, this corresponds to the global crisis. In this period, the average volatility was 35.4%, with a peak value of about 70%. This value of 70% reminds us of the peak value seen with the S&P 500 implied vol, in October 2008, where the VIX touched 87% intra-day.

August 2009 to June 2012: The market returned to an average IV of 21.9%, which is quite close to the value seen in the previous period.

June 2012 to May 2018: The final period is one of calm, starting from Jun 2012 onwards, where IV has averaged a level of 15.2%.


The IV is the market's perception of future volatility. It is interesting to think about what changed in the economy that changed the views of financial market participants across each of these break dates. Why was the IV low in the first period, then high, then low again? Why is this latest period the lowest?

This was the period of demonetisation and Donald Trump. The world is beset with economic and geopolitical risk. In India, the trailing P/E has risen to historically high levels. There is a banking crisis and a large fraction of the equity index is banks. The market seems to have conquered these fears and thinks that the future Nifty volatility will be low. This is not only true for Indian equity: the US VIX also achieved its lowest values ever (about 9%) in November 2017.

It is interesting to look back at previous periods of calm and of stress. For instance, we see that in 2007, implied vols were unusually low. Once again, this was similar to the behaviour of the US VIX before the crisis as well. There are concerns about the extent to which market participants are good at thinking about one security at a time vs. being good at macro forecasting.

How to decontrol the price of oil

We know a lot about price controls from the field of exchange rates. Here's an argument from way back, in 1998:

When change comes to a stabilised currency, as it must, that change is painful. Change in the long term is inevitable. The random walk doles out a little change every day, which is less painful than sudden large changes. 

...

Currencies which are random walks yield a deeper sort of stability. The steady pace of small changes every day generates realistic expectations about currency risk and continual realignment in production processes in the economy. It avoids sudden changes, and keeps the currency out of the domain of politics. The random walk regime is sustainable without incurring serious distortions in the economy.

In the field of exchange rates, India understood these arguments, and moved to a floating exchange rate. In March 2007, the INR/USD volatility moved up to roughly 9% and from early 2009 onwards, RBI stopped trading in the currency market. This was the biggest achievement of the UPA in economic reforms: In the 2007-2009 period, we got to a market determined rate on the most important price of the economy.

These same ideas are useful in thinking about the price of petrol. A large jump of Rs.1.8 per litre attracts attention. It is far better to let the price fluctuate every day. Ultimately, the price has to adjust. We suffer a lower political cost by letting it adjust every day (through the depoliticised market process). If we bottle up the small changes, then we have to make large changes. These are a bad use of political capital.

Jittery regimes fix prices

The puzzle

All of us are now curiously thinking about the abrupt phase transition that seems to sometimes occur in the endgame of an authoritarian regime. The traditional script was: The people rise up to rebel and the strongman murders them.

When the USSR collapsed, we thought it was special: it was a defunct regime that had just lost the will to live. But for the rest, the basic rulebook stood: the people get mowed down. And sure enough, that happened in Tiananmen Square.

But now, there are an increasing number of success stories with `velvet revolutions', and one has to think more carefully about what goes in an authoritarian regime.

Conflicts beneath the surface

What appears like a monolithic regime from the outside can actually often reflect a diverse array of interests tugging in different directions. In this beautiful article by Laurence Wright on Saudi Arabia, he says:

I had begun to look at Saudi society as a collection of opposing forces: the liberals against the religious conservatives, the royal family versus democratic reformers, the unemployed against the expats, the old against the young, men against women.

On that same thread, Why do protests bring down regimes? A follow up by Graeme Robertson says:

While the news media focus on "the dictator", almost all authoritarian regimes are really coalitions involving a range of players with different resources, including incumbent politicians but also other elites like businessmen, bureaucrats, leaders of mass organizations like labor unions and political parties, and, of course, specialists in coercion like the military or the security forces. These elites are pivotal in deciding the fate of the regime and as long as they continue to ally themselves with the incumbent leadership, the regime is likely to remain stable. By contrast, when these elites split and some defect and decide to throw in their lot with the opposition, then the incumbents are in danger.
So where do protests come in? The problem is that in authoritarian regimes there are few sources of reliable information that can help these pivotal elites decide whom to back. Restrictions on media freedom and civil and political rights limit the amount and quality of information that is available on both the incumbents and the opposition. Moreover, the powerful incentives to pay lip service to incumbent rulers make it hard to know what to make of what information there is.

I have also read others write similarly about China (but sadly, I do not have the reference): That in the absence of freedom of speech, the regime actually has no idea about where the problems lie, and is hence hypersensitive about criticism, and about solving the problems that it thinks do matter.

The behaviour of a jittery regime

Democracy matters in two ways. First, the regime has legitimacy. It is not worrying about a sudden upheaval that will destroy the regime. And, freedom of speech carries a steady flow of information to the regime. The UPA leadership does live in a bubble, but even they know that 8% inflation is a serious problem.

When a regime lacks legitimacy, and does not know what is going on, it is constantly fearful. It does not know what is going wrong and it can go off into extremes in trying to stave off some problems that it believes are first order. One area where this shows up is inflexible prices. To an external observer, it may be obvious that allowing price flexibility is better, but the regime is terrified about what will happen, so the price stays fixed.

Three examples

Egypt

In a blog post titled Garam Masala: Bread And The Life Of Egypt, Vikram Doctor writes:

I first realised how different Egypt was when I saw the bread in the street in Cairo. It was piled on low charpoy-like tables, thick rounds of freshlybaked bread, slightly scorched from the oven, a bit like tandoori rotis, but heavier.... Someone would replenish them from the bakery close by, and collect the money that people left, but nothing seemed to stop them just taking it away... the other reason why no one took the bread free was that it was so ridiculously cheap that they might as well just leave the few coins needed (in fact, buying bread seemed to be pretty much all that the piastre coins were used for). I calculated that, at that time (over 12 years back), the cost of a round of bread converted to something like three paise : something I could not imagine anything costing in any large Indian city. But this was the point: the price was unreal because a massive bread subsidy was one of the basic ways the Mubarak regime stayed in place.

Iran

From The regime tightens its belt and its first, in the Economist:

From top ayatollahs to the IMF, everyone agrees that spending $100 billion each year to pin down petrol, gas and electricity prices, besides the cost of staples such as flour and cooking oil, is a bad way to dispose of Iran's hydrocarbon revenues, accounting for more than 10% of GDP and encouraging waste on an epic scale. The symptoms of the malaise are legion: tea kettles simmer all day; the streets clog with recreational drivers out for a spin; lights glare because no one can be bothered to turn them off. `We can do it because we have oil,' Iranians used to tell incredulous visitors.

China

The outstanding price inflexibility of China is that of the exchange rate. Consider the Chinese and the Indian exchange rates of recent years:

There is a dramatic difference in the exchange rate flexibility. The Chinese authorities are extremely loath to allow the exchange rate to fluctuate, even though it induces massive distortions in the economy. Why? I would venture to guess that once a large export reprocessing sector has built up, the regime is just scared to rock the boat, to displease many workers.

The exchange rate is the most important price in any economy. A country that can handle a floating exchange rate is a flexible economy, one in which firms are born and die, workers move across locations and industries, and prices fluctuate. Deep and liquid markets are shock absorbers. Firms have ample equity capital, i.e. low leverage, so that they are able to absorb shocks. There is a whole configuration of institutional arrangements which are conducive to price flexibility. By and large, India fares well on these counts, particularly in the vast informal sector where there is extreme flexibility. And most of all, when things do hurt, individuals are able to express their discontent through democratic politics.

If India did not have these long-standing strengths, Governors Reddy and Subbarao would not have been able to move to a flexible exchange rate. And this exchange rate flexibility, in turn, enables an array of other economic reforms in favour of a market-based system.

Also see: The message for China from Tahrir Square by Minxin Pei in the Financial Times and The Secret Politburo Meeting Behind China's New Democracy Crackdown by Perry Link, on the New York Review of Books Blog.

Stability that is illusory

The regime change of recent years should make us think afresh about the notion of `political stability'. Democracy is always messy: demonstrations, machinations of party politics out in the open, colourful and often intemperate figures on television, elections, change in the ruling arrangement. But at a deeper level, this can be a more stable arrangement; there is no revolution at the end of the tunnel.

Similar reasoning applies in economics. Economists have always known that when prices appear to be stable, they often mask real trouble underneath. It is far better to have a small fluctuation every day, i.e. a steady flow of vol. The alternative -- of clamping down on price movements on most ordinary days -- merely yields big price movements on some days, which are far more difficult to handle.

Economic agents are not fooled by this stability on the surface. As Mark Roe says on Project Syndicate:

Even if all of the rules for finance are right, few will part with their money if they fear that an unfavorable regime change might occur during the lifetime of their investment.
More importantly, the grim stability of the type displayed by Hosni Mubarak's Egypt is oftentimes insufficient for genuine financial development. Authoritarian regimes, especially those with severe income and wealth inequality, inherently create a risk of arbitrariness, unpredictability, and instability. They are themselves arbitrary. And everyone knows that beneath the stability of the moment lurk explosive forces that can change the regime and devalue huge investments. Because financiers and savers have limited confidence in the future, such regimes can't readily build and maintain strong foundations for financial development.

Implications

This is a `capitalism and freedom' style argument: that democracy and markets interact in the double helix of modern civilisation.

Price flexibility works best when there is price flexibility in a lot of markets. If all prices were fixed, and you only freed up one, then it could easily make things worse. It is hard, crossing the hump, and reaching over to the other side where all prices are flexible. And, price flexibility goes well with democracy. Flexible prices are constantly disruptive. Every day, there are a few pockets of the economy that are really getting hurt in the creative destruction. It requires a confident regime to take these fluctuations in its stride. A jittery and illegitimate regime may be more likely to clamp down on price fluctuations since it fears these could destabilise it.

What could go wrong if petroleum product prices are decontrolled?

I have an article in Financial Express on this today.

Understanding food prices

I wrote an article in Business Standard titled Understanding food prices.

Paper on forecasting nifty volatility

Vipul and Joshy Jacob of IIM Lucknow have done a paper Forecasting performance of extreme-value volatility estimators in Jnl Futures Markets. As far as I can tell, there is no freely accessible PDF file of this article on the web. The abstract reads: This study evaluates the forecasting performance of extreme-value volatility estimators for the equity-based Nifty Index using two-scale realized volatility. This benchmark mitigates the effect of microstructure noise in the realized volatility. Extreme-value estimates with relatively simple forecasting methods provide substantially better short-term and long-term forecasts, compared to historical volatility. The higher efficiency of extreme-value estimators is primarily responsible for this improvement. The extent of possible improvement in forecasts is likely to be economically significant for applications like options pricing. By including extreme value estimators, the forecasting performance of generalized autoregressive conditional heteroscedasticity (GARCH) can also be improved.

What changed about the rupee: level vs. volatility effects

Business Standard has an editorial Rupee: level vs. volatility, which distinguishes between the two things that have happened on the Indian rupee: a shift in the level and a shift in the volatility:

The new currency regime that the RBI is following has led to new difficulties for individuals and firms. Two things have happened at once: The level of the rupee/US dollar has shifted from Rs 45 to Rs 41 per dollar, and the volatility of the exchange rate has gone up. The first effect inflicts pain on some (exporters) and pleasure for others (importers and the broad population). The second effect inflicts pain on importers, exporters and financial firms. The unhappiness about the shift in the exchange rate is not surprising. When diesel prices shift to their free market level, diesel consumers will obviously not be pleased. However, it is increasingly clear that it makes sense for monetary policy to focus on inflation control and not on targeting the exchange rate.

An entirely distinct story is the shift in volatility. The new currency regime represents a qualitative change from an RBI-controlled rate to a market determined exchange rate. The footprint of currency risk runs far beyond direct imports and exports. For, a large number of commodities are now priced by import parity pricing, where the local price is just the exchange rate multiplied by the world price, even if the producer and consumer are both Indian. Importers, exporters, financial firms and households are taken aback at this upsurge in volatility. This concern is legitimate.

The existing currency forward market does not pass muster, for numerous reasons. It is a non-transparent market, where substantial fees are transferred from users to financial firms. The non-transparency induces poor price discovery, something that India can ill afford at a time when well-functioning spot and derivatives markets on the currency are the need of the hour. The opacity of the existing forward market is a recipe for scandal, and this fear of scandal will continually generate a bias in the government to prevent this market from gaining adequate liquidity. Currency risk is present in every corner of the country, while the currency forward market takes place only in south Mumbai. There is a clear way out which addresses all these problems: that of establishing transparent, exchange-traded markets for both spot and derivatives on the currency, so as to match the market quality of the equity market.

In an ideal world, the finance ministry should have better planned the transition of the currency regime. First, the currency spot and derivatives market should have been built up, and only after these structures for trading and insurance were in place, currency volatility should have been permitted to go up. Such planning has, alas, not been done. It is, yet, not too late to urgently move on establishing new market structures through which individuals and firms can do better risk management. The Mumbai International Financial Centre report has proposed two key initiatives in strengthening the currency market: Establishment of a currency spot market, with a market lot of Rs 1 crore, accessible to all financial firms, and establishment of a rupee-settled currency futures market, accessible to all. The ministry needs to make up for lost time by rapidly executing on these two ideas, harnessing the strengths of Sebi, the NSE and the BSE. Rupee volatility like that seen in mature market economies must be accompanied by the tools found in mature market economies for coping with currency volatility.

I have previously written about currency futures, and there is a page on this blog about the MIFC report.

We live in interesting times

Just when you thought financial markets had become boring, things got interesting again.

• First, the Chinese market dropped 9% [picture]. That wouldn't, by itself, worry me so much. The stock market isn't that important to China's economy. It isn't a particularly well done stock market either, so I don't worry so much about the information content of this drop in prices. You might find this book review to be mildly interesting. William Pesek has a good column on Bloomberg, and this IHT story describes recent events.
• The big news is that last night, while we were sleeping in Indian Standard Time, the S&P 500 dropped 3.47% [picture]. Menzie Chinn has a great blog post diagnosing what happened, Caroline Baum has a good column on Bloomberg about the US housing market, and this NYT story uses the R word. This one-day drop of 3.47% is a fairly big move for the S&P 500 which has a daily sigma of 1%. If I look at the post-1990 period, on 99% of the days, the one-day returns were milder than -2.6%. They seem to get three days in each four years, on average, with a move of worse than -3.47%, and the last four years have been unusually benign.
• Most world indexes have also dropped [link]. I think this is mostly in response to the drop in the S&P 500 and its changed vol, and not so much in response to events in China.
• For many months now, the volatility of the S&P 500, as forecasted by the index options market, has been slumbering at remarkably low levels. It has jumped back dramatically [picture], going from near 10% annualised vol on Tuesday to roughly 18% last night. I wrote about the eerie low volatility a few weeks ago, and many commentators have been worried about the extent to which assets are `priced to perfection'.
• Turning to India, Business Standard has an excellent edit showing you the political context of this budget speech.

It's now 9:40 and at 11 the budget speech will start. This is, by itself, a big dose of vol for the Indian equity market. There could not have been a more dramatic setting for the speech. In the past few days, Nifty's been a bit volatile, so margin levels are already much higher than they were just ten days ago.

Beware the tranquility

Global equity volatility and Indian equity volatility is low. The blue line is the implied volatility off the S&P 500 options market; the red line is the backward looking rolling window volatility of Nifty (width = 66 days).

I wrote an article in Business Standard titled Beware the tranquility. In this, I worry that the world could undergo a `phase transition' and jump back to a configuration of high volatility, high risk premia and high liquidity premia. This is linked to the arguments made by Raghuram Rajan in June 2006 where he draws a link between low interest rates and the remarkable decline in risk premia and liquidity premia.

Larry Summers wrote a somewhat similar cautionary piece on 26 December titled Lack of fear gives cause for concern. Kenneth Rogoff discusses possibilities for the equity premium on Project Syndicate. And if you like to see these things on the sweep of history, look back at the First World War.

Volatility of the Indian equity market

The mass media often has a conspiracy theory view of stock market volatility. E.g. see this piece by Sucheta Dalal. The empirical evidence supports more prosaic interpretations. I did a talk at ICRIER on the subject of Indian stock market volatility. The PDF file of the slideshow might interest you.

Flying blind

While futures and forwards are merely priced by arbitrage, options trading is fascinating in that it throws up new information in the form of the implied volatility, which is the market's view about future volatility. This is something which is not visible before we have options trading but is visible after options trading commences.

In addition to revealing what private agents believe, it is also an excellent free forecasting tool. By now, there is a rich literature which tells us that such market-based measures are forward looking, and are very good forecasters.

In several other situations, financial markets are shaping up as new sources of forward-looking information for the statistical system: this includes market-based measures of inflation expectations, market-based measures of the probability distribution of the decisions of the US Fed in the next two meetings, etc.

Traditionally, it is felt that the Indian license-permit raj in finance induces real costs by virtue of supressing information about the yield curve and the exchange rate. In an article in Business Standard today, I argue that our approach of financial repression induces costs which go beyond the lack of a trusted yield curve and the lack of a trusted exchange rate. The host of information sources which are based on derivatives markets are also lost when policy makers block the development of financial markets. These information sources play a marvellous role in the decision making of private agents and of public policy in mature market economies, but we don't have them in India.

I worry about the illusion of control that goes along with flying blind. In the present situation, we don't observe a market exchange rate or a market yield curve, we don't observe market-based estimates of future inflation and inflation volatility, we don't observe market-based estimates of future currency volatility, and so on. I think that in a developing country, where the economists and the econometric models are weak, the benefits from having market-based forecasts is even stronger than is the case in industrial countries. (Of course, this goes over and beyond the minor matter of not having a decent measure of inflation).

Update: Jayanth Varma has a nice post on the delicious exchange-traded binary option on the Fed Funds rate that's gaining momentum at CBOT.

Coincidentally, today's Business Standard also carried a debate between Bibek Debroy and myself on the subject of capital account convertibility.

Talk by Raghu on monetary policy & asset prices

Raghuram Rajan has done a wonderful talk linking up agency problems in fund management with monetary policy. I have sketched a quick summary of his argument here. But the talk (~ 3000 words) is wonderful and well worth reading.

The literature has long emphasised the difficulties of the principal-agent problem between fund managers and their customers. Index funds only cost 1 bps to manage; so the only justification for almost all of fees & expenses is alpha. But there are only five sources of alpha:

• To be a Warren Buffet and identify undervalued assets.
• To create value out of activism - whether a hostile takeover of a poorly managed/governed company, or private equity investment.
• Financial engineering - innovate with creating new kinds of cashflow-streams.
• Liquidity provision.
• Sell deep out of the money puts, to generate an appearance of a steady stream of income. For many years, this will look nice, and then occasionally things will blow up. Raghu calls this "tail risk seeking" behaviour.

The first is truly hard. The 2nd and 3rd are feasible but highly competitive. That leaves the fourth and the fifth that are an ideal field of play for ordinary fund managers. A great deal of "active fund management" is really about earning alpha as a fee for giving out liquidity services or giving out tail risk insurance. (This argument is generic, and holds whether r_f is high or low).

Even if you are a smart fund manager, your customers can be stupid. Customers seem to do stupid things, taking assets-under-management (AUM) away from poorly performing managers and giving them to managers who have recently produced high returns. They thus generate wrong incentives for managers, and send AUM towards managers who exhibit these kinds of behaviours.

How does this situation link up to interest rates? Suppose interest rates are low. A finance company that has put out an assured return product promising (say) 6% returns is hard-pressed to produce 6% returns when r_f is low. It gets pushed into high risk assets. Similarly, the 2+20 hedge-fund compensation structure delivers higher absolute compensation when r_f is high. When r_f drops, the manager has an incentive to take bigger risks (or go into illiquid assets) to hang on to old levels of compensation. This is particularly the case when the compensation for the manager starts after a `hurdle rate' of an absolute level of returns is attained. When r_f is lower, expected returns on all assets are lower, and the hurdle rate is harder to reach.

In an ideal world, shifts towards high risk and low liquidity assets by some speculators should be compensated by other rational investors - e.g. a person trading with his own money who does not get into the issues of principal-agent problems of fund managers. But in a world where the bulk of liquid wealth of the planet is managed by external fund managers, institutional investors are large compared with rational speculators. Aberrations in their behaviour can then generate systematic distortions in asset prices.

Raghu's conjecture is that when r_f is low, agency problems generate incentives for institutional investors to go into illiquid assets, high risk assets, and returns from tail risk. The size of these investors is big enough that on the scale of the world economy, it looks like there is "heightened risk tolerance" - e.g. the very low values of the VIX when Greenspan had low interest rates. These flows are reversed when r_f becomes higher. Raghu cites Kashiwase & Kodres who find relationships between the VIX and emerging market spreads also.

In addition to traditional notions of monetary transmission, such effects constitute an additional channel through which monetary policy impacts upon the economy (though disentangling the channels of influence will be hard). This could have many implications for our understanding of how monetary policy works, and how optimal monetary policy should be crafted.

From an emerging markets perspective, this could be particularly important, because emerging markets offer risky and illiquid assets. I think it further undermines the case for a pegged exchange rate in India, for autonomous monetary policy would be a precious thing to have when faced with capital flows and trade flows which are strongly correlated with the world business cycle, and a local fiscal policy that can't stabilise. The only thing that can stabilise is local monetary policy, but there is a need for greater autonomy of local monetary policy.

Lessons from recent market volatility for the margin system

Margins are good faith deposits posted by market participants. While margins are normally defined by a rule book, recently there were two purely ad-hoc changes to margin required by SEBI. SEBI did an ad-hoc increase in one component of margins on 8 April, before the present bout of market volatility began. SEBI then did another ad-hoc decrease in margins on 25 May, while this bout of volatility was still underway. It is important to think carefully about the merits of such discretionary changes.

What should the correct margin be? The correct margin on a position is the size of the loss on this position that will be "rarely" exceeded. Jayanth Varma's Risk Management Group focused on a margin system which requires the sum of Value at Risk (VaR) at a 99% level, and the ETL or "Expected Tail Loss" which is E(r | r < VaR)), under certain simplifying assumptions. Intuitively, there is one component of the margin which is the VaR, which takes care of a loss on most days, and then on a few days, the loss is bigger than the VaR, but the ETL takes care of the average loss on those days.

VaR estimation is done adaptively in these schemes. Financial volatility is reasonably predictable: volatile days tend to be followed by volatile days, and vice versa. An adaptive system of margins involves charging low margins for the normal sleepy days, and driving up margins when higher volatility shows up.

So margins are raised after volatile days and vice versa. When market volatility goes up, participants are forced to put up more capital to support positions. I think this is an unhappy but essential feature of a sensible margin system. The alternative is to charge high margins all the time - which wastes capital. As long as margin changes are purely rule driven, market participants have correct expectations about what margins will be charged, under what circumstances.

Interestingly enough, the events of May 2006 were easier to handle, for the risk containment system, as compared with May 2004. Look at Nifty returns in both months:

Date	2004	2006
4	1.48	0.39
5	0.93	0.43
6	1.26
7	-1.56
8		0.79
9		0.74
10	-1.98	0.90
11	-4.02	-1.43
12	0.68	-1.39
13	0.37
14	-8.19
15		-4.11
16		0.58
17	-13.05	3.12
18	7.97	-7.01
19	4.16	-4.28
20	-1.54
21	1.05
22		-5.23
23		3.76
24	3.07	-2.65
25	-0.13	1.98
26	-0.49	1.00
27	-0.78
28	-5.02
29		0.16
30		-0.92
31	-1.68	-3.65

In 2004, the first hint of higher vol was the drop of 4.02% on 11th, which drove up margins a bit. Then, out of the blue, came the 8.19% drop on 14th (which drove up margins further). This was good preparation for 17th, which was a huge change. In contrast, in 2006, the 15th was the first big move, of 4.11%, which drove up margins. After that, margins were higher, and there was no real challenge from large price movements.

There are many myths about margins. One view is that increasing margins "cools the market" and pushes down prices. Conversely, it is felt that reducing margins tends to help "prop up the market". However, higher margins hurt both buyers and sellers equally! There is no simple relationship which asserts that higher margins yield lower stock prices, and vice versa.

The more subtle relationship is one where higher margins make it difficult to hold positions, thus reducing market liquidity. Indirectly, one could get a liquidity premium story whereby higher margins drive down liquidity and thus drive down prices. The biggest challenge at a time of market stress is liquidity: what we need most is that participants do not panic and retreat from trading. From a public policy perspective, increasing margins and thus reducing market liquidity - at a time when liquidity is needed most - doesn't seem like a bright idea.

Some believe that the system of margins induces a spiral of selling where a person is forced to make good his losses, and simultaneously submit bigger deposits, and thus collapses into distress selling. This view is inconsistent with the fact that derivatives trading is a zero sum game. For each speculator who has lost money, there is an equal and opposite speculator who has made money. While half the participants feel pain, the other half are feasting in huge profits. Just think of the joy of those who were short Nifty in the hours when Nifty dropped - they made huge profits. At the level of the country, these effects cancel out.

There are no permanent longs and there are no permanent shorts. The people who happened to be short at the right time got a lot of cash, and it is perfectly feasible for them to flip around the next instant and become buyers, if their speculative view changes.

Margin systems in India are imperfect. There is certainly more work to be done on improving the system of margins. The areas for work lie in:

• Better handling of liquidity risk,
• Shifting away from the simplifications of SPAN and RiskMetrics, and
• Shifting towards portfolio margining.

However, the basic logic of the margin system is sound, and there is little doubt in my mind that the system is strong enough to deliver soundness in the face of the time-series of Nifty returns. As evidence, note that these very systems were fine in coping with the more-daunting events of May 2004 - at which time no discretionary or ad-hoc margin changes were in the play. By international standards, we hold too much collateral. If there is a flaw in the system of margins, it lies in charging too-high margins in coping with model risk. Addressing the above three problems will help in reducing model risk and thus the extent of over-margining.

In this setting, I am unimpressed by SEBI's discretionary margin-changing decisions. Margins were raised, which hurt market liquidity, at a time when liquidity was needed most. Margins were reduced at a time when market volatility had not yet subsided. Now that SEBI has embarked on discretionary margin changes - which was not done earlier - SEBI's margin-changes are now an additional source of uncertainty for the market.

Regulators in mature market economies do not engage in discretionary margin changes. The job of the regulator is to think about the rules of the margin system. If there are problems with the rule-book, SEBI should get involved in changing the rule-book. But the job of SEBI is not to step in and make ad-hoc, discretionary changes.

In Latin America, employees of central banks and financial regulators are known to trade on currency and equity markets in order to profit from inside knowledge about government actions. Before any such accusations develop in India, it is better to emphasise a purely rules based system.