New risk management measures

By Fuzail Zubaid Ahmad

The stock market went down last week, but not due to changes in the margins regime that seems to have taken the back seat. This article is about the new risk management measures, which are likely to haunt the market after some time if not in the immediate future.

Members of stock exchanges, especially those who target the retail clients and generally handle highly speculative trades fear that the new Risk Management System (RMS) will ask for a much higher exposure margin that cannot be passed on to their clients. The new RMS asks for a primary and a special margin, both of which are mathematically generated figures, depending on a share’s historical liquidity and price volatility, as well as the latest spurt or drop in prices. Apart from that, the new haircut regime will also likely reduce substantially a share’s value acceptable as deposit with the stock exchanges. The mock exercise that is supposed to be underway at the stock exchanges will make it clear whether these fears about substantial rise in exposure margins as well as haircut are justified or not. If these fears are correct then SECP needs to implement the new RMS gradually.

VaR based margining system

Exposure Margins could be either plain vanilla or more real world (or Value at Risk) based. Plain vanilla means that all shares are treated in the same category of risk, regardless of their liquidity and volatility. The largest market cap company of Pakistan OGDC went over Rs200 in early March 2005, and came down to Rs110 by end of the same month. Plain vanilla exposure margin on OGDC would remain the same, as they did throughout March 2005, as plain vanilla exposure margining system was in vogue. A Value at Risk (VaR) based margin is more sophisticated in that it sets apart shares on the basis of their trading risk (volatility and liquidity) and then determines exposure margin for each share. Evidence has shown that Pakistani stock market is a risky place; hence a VaR based margining system would be more appropriate as it substantially raises the exposure margin in a risky price run on a share and vice versa.

Proof lies in eating the pudding. It would be a good exercise to test this VaR for the whole March 2005, and see what would have been the exposure margins in someone who had taken positions exclusively in OGDC, PPL, POL, NBP and MCB or a combination herein.

No VaR model developer claims that model to be the best. Models are mathematical in nature, but model selection is subjective, which is why Models range from being overly conservative to overly aggressive. The details provided by SECP about the VaR model it has selected for stock market (NCEL Riskmeter) are not enough, and members have a right to know the rise and fall in exposure margin requirements with real-life changes in share prices, to be able to tell if the model was too conservative or just about okay.

The new VaR based margin system puts companies in 6 different risk categories. If today were May 15, 2006, the exposure required for least risky category (A) shares would be only 15 per cent, 18 per cent for the next risk category, then 22 per cent, 28 per cent, and 43 per cent for the next three and finally 100 per cent for the most risky category of shares. These exposures for different risk categories would be obviously different on different dates because of different prices. Interestingly, on May 15, 2006 that was Monday, OGDC closed at Rs152.50 and at this price it was in category B, inviting 18 per cent exposure margin on that day. At Rs66.30 closing price, CTTL would have invited 43 per cent exposure margin as it was in the E category, only one short of the most risky category F.

Special margins

Some people also question the wisdom of special margins. New RMS asks for special margins in shares whose prices have risen or fallen too sharply. This shows that abrupt price movements are not adequately captured by NCEL Riskmeter to change the exposure margins adequately. NCEL Riskmeter experts need to comment in detail.

Secondly, when a particular share undergoes a large and sudden price fall, the sellers have to deposit the special margin. In this case, considering a trailing PE multiple of 15 seems too large because it will be obviously higher than the 26 weeks moving average price of that share and the seller will have to deposit special margin to address the difference between a low sale price and a 15 PE based price.

Haircuts

Haircut indirectly serves as another special margin. The only criterion selected in the new system is the impact cost, as against the existing system of using turnover and EPS. The new RMS uses NCEL ICAnalyzer software to determine the haircut applied on the face value of shares for accepting them as eligible securities for depositing as exposure margins. This means that first NCEL Riskmeter will determine the primary exposure margin, then the system will generate special margin figure that may be collected from buyers or sellers in times of heavy price fluctuations, and then the NCEL ICAnalyzer will determine the haircut on eligible securities’ acceptable values as deposits. Maybe the new RMS is overly conservative, but it has to be tested dynamically in real-life situations, as mentioned above.

Key objectives of RMS

Raising exposure margins (by way of primary and special margins together with haircut) automatically means improving debt to equity ratio. Any rise in members’ exposure margins is supposed to indirectly reduce position leverages by the members’ clients, because higher margins are supposed to be collected from them. This means that all brokerage houses and individual members must ask their clients to deposit higher margins on their orders, so that these clients raise their equity portion in their holdings. In developed and many fast developing economies, exposure margins or simply margins are often raised to confront likelihoods of asset bubbles. So if someone believes that exposure margins are meant to reduce members’ default and settlement risk, he or she is missing a big part of the picture.

Under intense pressure from the members of the stock exchanges, aided by the falling market, SECP exempted financial institutions (with certain conditions) from depositing any margins. Financial institutions do not normally speculate, but this is true for other institutional investors also, so it makes a good case to exempt them too, provided they possess certain minimum credit rating.

Highly leveraged speculation is a tool almost exclusively used by the unsophisticated, uneducated, poorly informed individuals (called jobbers, day traders, punters, etc.). He is and should be the main focus of any margin raise. However, what the RMS needs to do is make sure that this individual speculator deposits his share of the margin (normal plus special). The members should not be able to use their own wealth (proprietary cash and shares) to deposit margins on behalf of these speculators because it will defeat the very purpose of new RMS.

Member defaults

There are occasional member defaults in the Karachi Stock Exchange, except that the matter is privately handled and settled by the members themselves and not reported to stock exchange management or the apex regulator. This private settlement of member defaults mostly results in changes in the shareholding structure of some brokerage houses of Karachi Stock Exchange changes. But why do members and brokerage-houses default? It is because the individual member or the major shareholder of a brokerage house takes an untenable position in one or more of the three markets (ready, futures and CFS) that is far beyond his capacities to sustain in times of adverse price movements. When he fears a default, he offers his entire or part shareholding in his card, and requests the buyer, who is usually a big member of KSE himself, to let him retain the name of his brokerage house, so his honor remains intact.

To relate this feature with the subject, these defaults occur because the existing RMS allows high-leveraged speculation in the stock market, but not the new RMS that seems far more conservative, but it has to be tested in real-time mode.

Conclusion

The mock exercise might tell us that the RMS is overly conservative, in that the margins requirement goes up steeply when a share goes up only slightly, and margins requirement does not fall significantly when the price falls. Or the mock exercise could tell us that the new RMS is behaving in a manner that seems entirely acceptable to all the members of KSE. This means that the mock exercise is going to be an important thing to do before a full implementation of new RMS is demanded by SECP.