Tag Archives: Corporate Life

Rates and Valuation

Marking trades to market requires up-to-date market data. There are two types of market data required for pricing — one is the live spot rates, volatilities, interest rates etc. This type of data is collectively called rates. The second type is the kind that goes into defining the products being traded, or the characteristics of the rates. These include definitions of interest rate pillars, bond coupon dates and rates etc. This second type is considered static data.

Valuation and Product Control

The rates management team is in charge of the first type data. They ensure that the live data providers are consistent with each other and that the data itself is accurate. They do this by applying various automated tests and limits to the incoming rates to flag any suspicious movement or inconsistency. Once approved by the team, the data gets consumed by the trading platform. The rates management is a critical role, and the market data is often stored and served in dedicated databases and services. Because of the technicalities involved, this team works closely with the information technology professionals.

The static data is typically managed by a separate team independent of rates management. They go by various names, Treasury Control being one of them. They set up traded products and rates pillars and so on. In some banks, they may also be responsible for trade input data validation.

Two other important functions of Middle Office are valuation and product controls. These functions are pretty far removed from quantitative development and trading platform. These teams ensure that the trade valuations and P/L movements are consistent with market movements. Valuation Control takes a close look at pricing and P/L mostly at trade level while Product Control worries about P/L explanation typically at portfolio level. Since we have the Greeks (rates of change of product prices with respect to market quantities and time), we can compute and predict the change in the prices (or P/L movements) using Taylor series expansion. If the independently computed prices (using actual market rates) are at odds with the predicted ones, it points to an internal inconsistency and should trigger a detailed investigation.

Product Control may also help Finance and Human Resource with valuation reserves process, which estimates the level of exaggeration in the profit expectations of ebullient traders. Since traders’ compensation is tied to the profit they generate, this process of assigning reserves against profit is essential in ensuring equitable performance rewards.

Market Risk Management and Analytics

If you play in the market, you run the risk that it may move against you. This risk is, of course, market risk and we have a Middle Office team to manage it. Market Risk Management (MRM) ensures that the risk limits on the volumes and types of products traded are set in accordance with the risk appetite prescribed by the senior management. It also ensures, through regular processing and monitoring, that these limits are adhered to.

MRM

What is monitored are risk measures such as the Greeks and Value at Risk (VaR). The Greeks are the first and second order derivatives of the price of a security with respect to various market variables such as the price of the underlying, interest rates, volatility as well as trade specific entities like the time to maturity. The VaR is a statistical end point measure estimating the amount of loss at a given confidence level in the case of an adverse market movements, and is typically computed using the historical market movements over the past year or so. These risk measures are aggregated, sliced and diced in various ways to make it easy to monitor them, and reported to senior management, risk control committees, trading desks etc. The MRM team is also responsible for reporting to regulatory agencies, both in the form of regular compliance reports as well as ad hoc reports in response to drastic market moves.

Quants can find opportunities in the Analytics team embedded within MRM. This team is in charge of pricing model validation, which is the process of ensuring that the mathematical models deployed in trading systems and other valuations engines are both appropriate and correctly implemented. There is a significant overlap between the work that MRM analytics quants do and their Front Office counter parts (whom we called pricing or model quants). The Analytics team also takes care of any other quantitative tools needed in MRM or risk management in general. Such tools could include potential future exposures (PFE) for credit risk management, liquidity modelling for Assets and Liability (AML) etc.

Credit Risk Management

Risk management is a critical function of Middle Office. Credit risk is the risk that somebody who owes you money may not be able or willing to honor their obligation. In other words, they may default on their credit obligation. This risk is managed in a bank using a variety of statistical tools.

Middle Office

When a bank issues you a credit card, it takes on credit risk that you may not pay up. You pay an insanely high interest rate on your outstanding balance precisely because of this credit risk. The risk is not secured. A mortgage or an auto loan, on the other hand, is secured by the equity of your property, and you pay a significantly lower interest because of the collateral.

The Middle Office team of Credit Risk Management (CRM) operates using the same two paradigms. Much the same way as you have a credit limit on your credit card or line of credit, each counterparty that the bank trades with has a certain credit limit based on their credit rating as published by credit rating agencies such as Moody’s or Standard & Poor. The problem with this mode of managing credit risk is that the bank has no way of knowing how much credit is loaded against a counterparty’s rating in other banks. Nor does it have a means of finding out how many credit cards you have. In Singapore, the regulatory authority, MAS, tries to minimize the risk of people going bust be requiring that their credit limit be twice their monthly salary. Bt they may get as many credit cards as they want from different banks against the same limit, effectively nullifying the good intention behind the requirement.

This overloading against credit rating is avoided when the risk is managed using collaterals. Much like you cannot take two mortgage loans on the same property (not without adequate equity, any way), counterparties in trading also cannot use the same collateral for multiple trades. Banks and counterparties typically use bonds as collaterals and physically exchange them during secured transactions.

Before the Front Office trader can enter into a trading agreement with a counterparty, they will need to get approval from the credit controllers who will assess exposures and check them against predefined limits. The exposure assessment uses techniques such as potential future exposure (PFE) based on a large number of simulations of potential future markets.

In addition to the risk of counterparties defaulting during the life time of a trade, CRM professionals worry about the potential for default during the delay in settlement — after the maturity of a trade (where the bank is in the money) and its settlement. This risk is aptly called the settlement risk.

Middle Office

The structure of Middle Office in a typical bank is depicted in the slide below. The functional units within Middle Office work hand in hand with those in Front Office to handle the inception approvals and regular processing of trades.

Middle Office

Middle Office is different from Front OFfice in that it has little interaction with the external world. Its primary (and perhaps only) customers are the Front Office traders and teams. As usual, most of the interactions among the teams within Middle Office and Front Office take place via the trading platform, which acts like the boundary interface between the two Offices, as shown in the slide.

In later posts, we will go through the functions of each of the business units described as a box in the picture. For now, as a general summary, we can see that the Middle Office functions are of two kinds: those related to trade approvals based on projected risks and limits, and those related to regular trade monitoring. But these functions are vast in their scope, and require large systems, data flows and an army of professionals to carry them out. They are organized under the business units with names like Product Control, Trade Control (or Treasury or Business Control) Unit, Market, Credit and Operational Risk Management), Limits Monitoring, Rates Management, Compliance and Regulatory Reporting, Analytics, Asset and Liability Management etc. Again, keep in mind that this description of Middle Office is from the perspective of quantitative development relevant to structured products trading.

Quantitative Developers

If Quantitative Developers look like the heart of everything that goes on in the Front Office (according to the following slide, that is), there is a good reason for it. This series is written from the perspective of Quantitative Development. After all, the series, the talk, and the book are all titled “Principles of Quantitative Development.” From that vantage point, sure, we are at the center of the universe.

Quantitative Developers

To be fair, in structured products trading, quantitative development and quantitative mathematics play a crucial role. As we will see in later posts, almost all the aspects of trade lifecycle management are mediated by the end product of these quantitative professionals, which is the trading platform. Crucially, the trading platform defines the interface between Front Office and Middle Office. Within Front Office, quantitative developers act as the conduit of integrating the pricing models developed by quants into the platform, thereby making them accessible for profit making by trading desks. Because of this buffering role that the quantitative developers play, they have to field almost all of the support requests from trading desks and sales personnel in Front Office, as well as from anyone who uses the trading platform.

In the corporate organization, quantitative developers may find themselves under the information technology department, supporting the trading platform from afar. From a career perspective, this organization is less than ideal for them because IT is a cost center, not a profit generator and the compensation and remuneration schemes reflect that fact. Besides, IT tends to be considered as being outside the core business of the bank. Far better for them would be to find themselves embedded within the Front Office setting, where the quantitative developers can offer direct support to the stakeholders from within and enjoy the prominence and prestige that comes with the critical role of managing the vital in-house trading platform.

Average Beauty

If you have migrated multiple times in your life, you may have noticed a strange thing. The first time you end up in a new place, most people around you look positively weird. Ugly even. But slowly, after a year or two, you begin to find them more attractive. This effect is more pronounced if the places you are migrating from and to have different racial predominance. For example, if you migrate from the US to Japan, or from India to China. As usual, I have a theory about this strange phenomenon. Well, actually, it is more than a theory. Let me begin at the beginning.

About fifteen years ago, I visited a Japanese research institute that did all kinds of strange studies. One of the researchers there showed me his study on averaging facial features. For this study, he took a large number of Japanese faces, and averaged them (which meant he normalized the image size and orientation, digitally took the mean on a pixel-by-pixel basis). So he had an average Japanese male face and an average Japanese female face. He even created a set of hybrids by making linear combinations of the two with different weighting factors. He then showed the results to a large number of people and recorded their preference in terms of the attractiveness of the face. The strange thing was that the average face looked more pleasant and attractive to the Japanese eye than any one of the individual ones. In fact, the most attractive male face was the one that had a bit of female features in it. That is to say, it was the one with 90% average male and 10% average female (or some such combination, I don’t remember the exact weights).

The researcher went one step further, and created an average caucasian face as well. He then took the difference between that and an average Japanese face, and then superimposed the difference on an average face with exaggerated weights. The result was a grotesque caricature, which he postulated, was probably the way a Japanese person would see a caucasian for the first time.

This reminded me of the time when I visited my housemate’s farm in a small town in Pennsylvania – a town so small that the street in front the farm was named after him! I went with his parents to the local grocery store, and there was this little girl sitting in a shopping cart who went wide-eyed when she saw me. She couldn’t take her eyes off me after that. May be, seeing an Indian face for the first time in her life, she saw a similar caricature and got scared.

Anyway, my conjecture is that an averaging similar to what the Japanese researcher did happens in all of us when we migrate. First our minds see grotesque and exaggerated difference caricatures between the faces we encounter and the ones we were used to, in our previous land. Soon, our baseline average changes as we get more used to the faces around us. And the difference between what we see and our baseline ceases to be big, and we end up liking the faces more and more as they move progressively closer to the average, normal face.

Here are the average male and female faces by race or country. Notice how each one of them is a remarkably handsome or beautiful specimen. If you find some of them not so remarkable, you should move to that country and spend a few years there so that they also become remarkable! And, if you find the faces from a particular country especially attractive, with no prolonged expsosure to such faces, I would like to hear your thoughts. Please leave your comments.

  
[I couldn’t trace the original sources of these images. If you know them, please let me know — I would like to get copyright permissions and add attributions.]

There is more to this story than I outlined here. May be I will add my take on it as a comment below. However, the moral of the story is that if you consider yourself average, you are probably more attractive than you think you are. Than again, what do I know, I’m just an average guy. 🙂

If you found this post interesting, you may also enjoy:

  1. Why is seeing not quite believing?
  2. Sophistication

Trading Desks

At the heart of Front Office are the trading desks. In terms of prestige and power, they really are the reason for the whole infrastructure of Front Office, including economists, sales, structuring, quants, quantitative developers etc. After all, they make the profits. And consequently, the vocal and volatile traders hold enormous sway. At their beck and call, quantitative developers provide instant service on trading platform issues; quants develop pricing models based on their requirements.

Trading Desks

Trading desks interact with the external world of brokers and counterparties. They take their input on market moves from highly responsive market data providers and base their positional views on staff economists. They have an army of trading assistants (junior traders themselves) who help them book and monitor their trades with the help of risk management professionals associated with the desks.

Their interaction with the rest of the bank is mainly mediated by the trading platform. When the book a trade, for instance, it goes into the trading platform and ends up with some middle office professional who will decide whether to accept it or bounce it back for further modifications. Various risk management staff from the middle office also will take a hard look at the trade, as we shall see later in the trade lifecycle.

The desk risk management team get their cues also from the Middle Office risk management, in terms of approved limits and daily marked-to-market and sensitivity reports. All these channels of communication need to be facilitated in the trading platform.

Sales and Structuring

Those in sales tend to be personable, extroverted and persuasive people. A good salesman can sell a refrigerator to an eskimo, I’m told. IN the context of the front office in the global treasury or global markets, sales people are tasked with sniffing out trading needs from their customers and offer compelling products from the trading and structuring desks to fulfill them.

Sales

The life of a sales professional is a tough one. They need to meet progressively more aggressive targets in sales volumes to stay in the game. The moment they meet the target one year, it jumps the next year to an even more unattainable level. The sales staff, therefor, find themselves under constant (or even increasing) pressure. Knowledge of mathematical finance, while useful, is not a pre-requisite for sales jobs. In fact, mathematically inclined folks tend to be reserved and introverted, and tend not to have the qualities that make for a good salesman. In other words, if you are a quant, your most ideal role is not in sales, although the structuring side may offer some interesting opportunities.

Economists

Economists have too many hands. On the one hand, they may feel that oil prices are going to go up because of increased demand from emerging giants like India and China. On the other hand, there is a global slowdown, and the overall demand is likely to fall, putting downward pressure on the prices. Then again, the rampant corruption and inherent deficiencies in markets like India might push the prices high. On another hand, price-driven improvements on the supply-side (like better production techniques) may eventually push the prices down. My ex-boss, an economist himself, once told me that he wished he could chop off some of these hands!

Economists

But seriously, economists are the mouthpieces of the bank. When you see someone from an investment bank on TV making an intelligent observation about the market outlook, it is likely to be a staff economist. If you manage to navigate through the jumble of hands, you will hear what the bank wants to say to the world.

That is precisely why you have to pay close attention and try to read between the lines. You are listening to what the bank wants you to hear. When you listen to a Morgan-Stanley economist assert that Facebook is the best thing since sliced bread, should you trust him, given that they were the investment bank behind Facebook IPO? Of course, the lingo you will hear from the economist would be a lot nicer, subtler and more convincing than I can muster. But you should still ask yourself, “Is there a hidden agenda?”

Economists may seek to influence the market; they also distill and condense their take on the market and share it with the traders and executives so that they can formulate and fine-tune their tactics and strategies. Thus, they play an important role in the directional views that banks take when navigating the tricky waters of trading and speculating.

Front Office

The Front Office looks quite complex in the slide below. All we need to remember is that the “Front” in Front Office is for world-facing. So in the slide below, you can see functional teams, some of which interact with the external world. These teams are:

  • Economists who talk to the media and receive market inputs that they condense into useable intelligence for the rest of the bank.
  • Sales and Structuring teams who talk to customers to sniff out potential opportunities.
  • Trading Desks, interacting with their peers in other financial institutions and other counter parties.

Front Office

One level removed from this layer of customer or world-facing teams are the quants and quantitative developers of Front Office, along with the risk managing professionals associated with the desks. They play supporting roles to the first layer of teams. Quants develop pricing models, quantitative developers roll them out into trading platforms, and the desk risk management team helps traders monitor and hedge their exposures. Note that almost all the interactions among these teams are mediated by the trading platform. The platform also acts as a conduit between the functions of Front Office and Middle Office. The nature of these functions and interactions of each team will be the topic of the next few posts, starting with the economists. Stay tuned.