Search interesting materials

Saturday, June 03, 2006

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:

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.


  1. 1)Can regulators/exchanges ever control for market risk that is not captured by a VaR system?? There are, of course, a lot of criticisms about VaR and its usefulness in indicating statistical outliers..The most famous of which i have come across is that of Naseem Taleb:

    "VAR has made us replace about 2,500 years of market experience with a co-variance matrix that is still in its infancy. We made a tabula rasa of years of market lore that was picked up from trader to trader and crammed everything into a co-variance matrix. Why? So a management consultant or an unemployed electrical engineer can understand financial market risks.

    To me, VAR is charlatanism because it tries to estimate something that is not scientifically possible to estimate, namely the risks of rare events. It gives people misleading precision that could lead to the buildup of positions by hedgers. It lulls people to sleep. All that because there are financial stakes involved.

    To know the VAR you need the probabilities of events. To get the probabilities right you need to forecast volatility and correlations. I spent close to a decade and a half trying to guess volatility, the volatility of volatility, and correlations, and I sometimes shiver at the mere remembrance of my past miscalculations. Wounds from correlation matrices are still sore."

    2) The margin system that is enforced is for registered participants in the market like brokers. What about the margin that broker demand from their clients? It is natural to expect that the broker-client margin will be determined more by competition among brokers for clients rather than an extensive model based margin system (i.e. brokers who want to capture market share will let their clients run greater leveraged positions).. As it wouldn't be sensible for SEBI to ask brokers to maintain margins on this front as well, can a system of capital adequacy be specified so that only well capitalized brokers who can stand the risk of all their clients defaulting, be allowed to participate??

  2. Interesting paper on the daily volatility in the Indian markets..

    The study uses data that ends in Jan I guess it needs to be updated for more relevant conclusions..One of the few studies I have seen that uses high frequency data rather than end of day historical data...note the conclusion on FII activity...

  3. A clearing corporation must have a 100% rules-based system through which margins are determined. So in 1974, when SPAN was invented by CME, it was a big step forward, even though it's a pretty bad system based on our understanding today. Yes, 2,500 years of epxerience in the hands of human trader is great. But a clearing corporation requires rules, not discretion.

    I am the first to say VaR is not perfect, and that there is model risk. But I think it's better to setup a system using the best available models instead of trying to either inject human discretion into it, or of trying to fly by ordinary human intuition unaided by models. The unemployed electricial engineer's approach is a darn sight better than an innumerate approach.

    BTW, the present Indian situation assumes there is no benefit from diversification. I would be very happy if we could have a discussion about what's a better covariance matrix. The trouble is, right now there is none! That's the sort of thing I classify under "portfolio margining". The problem goes back to the CME SPAN system, which is pretty crude in thinking about the relationship between products.

    On the question of collateral charged by brokers of their customers, yes, it's very possible that there can be a race to the bottom where the broker who charges the least margin gains market share. But I think there's a fairly strong system of random inspections to catch and penalise such activity. My rough view is that it's hard to obtain nontrivial exposure in India today - as a client - even if the broker is a good friend of yours. That's one of the big achievements of the last 10 years.

  4. Hi Sir,

    Isn't a bank similar to a highly leveraged trader. If the CAR stands at say 10%, are the banks not leveraged to around 10-odd times. In comparison, a trader in the futures market pays around 20% (i think) to take positions. Leverage their is 5 times.

    What I really mean is can we a margin system in the financial markets much like the CAR for banks? Wouldn't life be simpler if someone told me that I gotta pay 15% as margins to take a position in stocks or commodities?

    In any case, we can never quite predict a LTCM catastrophe, or predict a nick leeson in the making. So why worry once you have a standard margin rule which is same for both bull markets or bear markets.


  5. As I have emphasised before, banks are hardly a role model for sound financial risk management! As Merton Miller has said, banking is a `disaster-prone 19th century industry' - a toxic mixture of too much leverage, non-transparent assets, clumsy regulation and moral hazard.

    With Basle-II there is an effort at taking away from dumb %-of-RWA type rules. Done right, this ought to take banks closer to a VaR type notion. Of course, model risk with banks is stupendous compared with what we see with securities.

    (You used the words bull market and bear market - these are wrong. Margins change with _volatility_ not the sign of returns).

  6. I like your article and it really gives an outstanding idea that is very helpful for all the people on web.


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.