StructuredDeals — Trader Construction Kit

Chapter 15 Excerpt - Pricing & Hedging Structured Transactions

Structured Transactions

Structured transactions can be found at both ends of the sophistication spectrum: seemingly simple products that contain difficult-to-disentangle organic risks, or obviously complex synthetic structures. Organic structured transactions frequently originate from naturals, who bring the risks that they would like to shed to banks and other financial intermediaries for pricing, hedging, or outright sale. At first glance the exposures can appear simple, but may contain complex risks arising from the physical attributes of the customer’s core business that could require a lengthy process of decomposition and pricing. Often there is no established methodology to value particularly esoteric products, with no historical data and no point of reference to use as a baseline. Naturals will also sell or lease the rights to entire facilities to merchants, particularly in early-stage markets, as they are the premiere optimizers of physical systems and have a real appetite for the exposures.

At the opposite end of the spectrum are structures born in a lab at a financial institution and intended for sale to customers as “risk management products.” The rise of exotic derivative markets in the 1990s exponentially increased the level of complexity of the products on offer by ushering in the practice of creating and pricing highly customized solutionsoften for problems end-users did not necessarily know they had. There is a price for everything, and in that intellectually adventurous and unfettered atmosphere risks were cut apart and sewn together in expensive and avant-garde combinations. The lack of standardization precluded evolved, well-vetted modeling, as practitioners could not justify the time to develop rigorous valuation methodologies for a product that may only ever trade once, if at all. The challenge for the prospective customer/victim of some of these esoteric products was the same one faced by the protagonist in every sci-fi movie who stumbles across the piece of bizarre alien technology: Who made it? What does it do? How does it work? What happens if I push the shiny, shiny red button?

A trader evaluating a structured transaction must attempt to re-model, reverse-engineer, or deconstruct and price the product based on the provided description or term sheet and his knowledge of the market space. The challenge is that the constructor almost invariably has superior information, greater familiarity with similar structures and their value, and better analytical resources. One unique aspect of structured transactions is that neither the buyer nor the seller generally has any interest in transacting near fair value, as each will want to build in a margin to protect from the risks he knows he cannot model and hedge. If the deal is worth X, the seller will only transact for X+ and buyer must buy for X- to allow room for sub-optimization, particularly for physical deals. Structured deals get done because the two counterparties have different information, different analytical and pricing capabilities, different forward projections, and different perceptions of value, all of which lead to different prices.

To handle all of this fancy deal flow a new species of hybrid employees called structurers evolved at both buying and selling institutions. A structurer has some of the commercial knowledge of the originators and traders combined with the analytical acumen of the quant and risk groups. The structuring group is tasked with examining complex deals, decomposing them into their component risks, routing each piece to a quant for valuation or a trader for pricing, then re-assembling the entire monster for presentation to the client or evaluation by management. In flowchart form the process looks like this:

Figure 15.1 Flowchart of the origination and structuring process.

The process frequently starts with a conversation between a customer (generally a natural) and the originator that covers it for a bank or financial institution. Most complex, highly structured deals begin with a customer inquiry, either in the form of a Request for Quotation (RFQ) or Request for Proposal (RFP).

An RFQ is a straightforward request for a market in a particular product broadcast to a range of counterparties, either a select pre-screened group or the market as a whole via an electronic platform. A natural might issue an RFQ for a volume of flat price directional hedges to a small group of banks, for example. An RFQ would typically bypass the structuring desk and go directly to the trading desk for pricing and execution. An RFQ request is also a green light for a good originator to attempt to up-sell the client on a more sophisticated risk mitigation strategy.

An RFP is a request for pricing of a more complex transaction, often one that is structured around the operational properties and constraints of an actual physical asset or production facility that cannot be cleanly hedged with standard, readily available products. For example, a hedger might issue an RFP to buy a structure that would let it lock in a portion of the current margins around a metals processing facility while allowing it to participate in a percentage of the upside price appreciation of the end-use product.

Modeling a Deal

The originator will take his notes or the customer’s RFP to the structuring desk to get a sense of the complexity of the valuation challenge, generally involving the trader with the most relevant experience. Evaluating proposals for structured transactions is a little like appraising the first cards dealt during a hand of poker. Most should be discarded immediately. Some are workable, and will usually be passed around for comment and possibly some light analysis to determine if there is anything worth doing. A rare few are clearly, unambiguously valuable and merit immediate attention and the expenditure of resources to evaluate and price.

If the proposal is interesting enough to price, the structuring group will begin to break it down into its component risks. For proposals similar to previously executed transactions, this can be a well-established exercise with the only minor wrinkles caused by the specific requirements of the customer. For truly novel pieces of business the structuring group will have to convene a small working group of analysts, a trader or two, and representatives from risk management in an attempt to intellectually crowd-source a solution to the problem. Typically the division of labor breaks down as:

• The structuring group will quarterback, making sure that it has the necessary information from each specialty and handling the more commercial modeling for the deal, including a framework for valuation that will source inputs from trading and quantitative analytics.

• The trader will source all market price, depth, and liquidity information, including live forward price quotes for futures, options, and spreads.

• The quant or analytics group will start to model the non-standard risks, which can require extensive simulation or the development of a valuation framework.

• The risk management group will ideally be present from the start to give the deal team insight into how the trade will be modeled in the deal capture system and the valuation challenges inherent in its mark to market. Very complex transactions that involve novel products and price locations may require the risk group to design new templates within the system to accurately model the nuances of the transaction.

• The credit and collateral group will have to compute the estimated funding requirements for booking the transaction and the associated hedges. The deal team will have to take this cost into consideration when pricing the transaction.

• Management will be brought in on any large, novel, or P&L impacting transactions, as they will ultimately have to approve the deal and its pricing before it is shipped to the client, and may have to proactively obtain approvals from senior management for any unusual risks or novel products/instruments inherent in the transaction.

The Trader’s Contribution – Pricing, Depth, and Availability of Products

Once the deal is broken down into component pieces, they will be shipped across to the trading group for pricing, either directly out of the market or via analogs to similar products. Frequently, traders will be expected to make markets in uncommon products or particularly illiquid securities. The implicit and sometimes explicit assumption is that the traders should be willing to take the exposure onto their books at the price they are quoting.

The trader will also be expected to advise the structuring group on the depth of the market to calibrate the assumptions of how much slippage and execution risk to build into the pricing. This is relevant on both an initial hedge basis (what will it cost to place a volume of hedges on Day 1) and as an ongoing optimization exercise (what will it cost going forward doing optimization-related transactions).

The trader will also comment on the reasonableness of the transaction assumptions built into the valuation and the associated hedge plan. It may be possible, given normal levels of liquidity, to expect to execute a large volume of option hedges at the benchmark product in the market. The quant group may utilize that benchmark option price to derive the price of a similar option in the market most applicable to hedging the transaction. It is the trader’s job to inform the quant and structuring group that, while its valuation makes intuitive sense, it does not guarantee that there will be counterparties willing to transact in the local market for any size, let alone the volume needed at the price imputed by its calculations.

Traders with an affinity for structured customer business become expert at understanding value differentials between standard benchmark products and illiquid, non-standard instruments, and develop a level of comfort compensating for uncertainly with price adjustments. The trader will start with what she knows and can observe, the product closest in performance to the puzzle piece she is trying to replicate, and then incrementally add or subtract value based on the characteristics of the product she is trying to simulate. When in doubt, the trader will err on the side of caution, as building in too much margin and missing the deal is preferable to building in too little and booking a loser.

Embedded Options

The trading group and the structuring group will both evaluate the transaction to ensure that all of the embedded optionality is discovered and accounted for. There are several varieties of embedded optionality.

There is straightforward economic optionality, where if a price or volume reaches an established threshold the holder of the option can exercise and modify the conditions of the deal as specified in the contract. This acts like a vanilla put or call submerged in the contractual language of the deal.

For example: A coal producer agrees to sell a power generator 1 million tons of coal at the prevailing price of $100 per ton, with contract language that increases the total size delivered to 2 million tons if the price drops to $95 prior to the start of delivery. It is relatively easy to see that this transaction is actually a 1M-ton sale with an associated 1M-ton $95 strike put option sold back to the coal producer by the power generator. The put is fairly straightforward to price, the only question being whether the power generator correctly valued the option and seeks an equivalent discount to the cost of entering into the contract.

Non-economic options can be highly difficult to recognize and all but impossible to model, particularly those having to do with operational and reliability characteristics of a physical asset. A need to bring a production facility down for maintenance will certainly impact a deal predicated on its ability to generate output, but at the same time may not be driven by or correlated with any particular economic signal.[72] A driveshaft or furnace does not consider the marginal cost of the output of the facility before snapping or tripping offline.

Once the deal team has identified all of the optionality present in the deal, there is the minor problem of figuring out how to model and price it. Recall the exploration of the standard Black-Scholes model in Chapter 10 and its many inherent structural limitations. While it is possible to bend the rules slightly and achieve reasonable results for plain vanilla calls and puts, most of the non-standard characteristics of a complex physical option would invalidate the underlying economic underpinnings of the model.

As a result, risk managers and high-level quants will assemble a collection of specialized options models to choose the most applicable tool to price the risk inherent in the transaction.[73] “Most applicable” is the key term, as some of the types of organic optionality commonly found in structured transactions cannot neatly be modeled by anything, forcing the firm to go with the closest thing in the arsenal and attempt to understand and live with the discrepancies, or to home brew a solution. Neither is optimal, and traders and managers will often have to spend significant time laying siege to Castle Math deep in the non-Euclidian forests of Quantsylvania in an attempt to decide whether they would prefer to go with something they know is slightly wrong (but hopefully only slightly) or something that could be completely right or completely wrong.

Recall the discussion of multiple internal valuation methodologies from Chapter 10; it is possible for different groups to value an option with different models and arrive at a completely different answer. This is extremely common with structured deals, and is another reason to have good coordination between structuring, the quantitative analysts, and risk from the beginning. The people doing the commercial valuation must know how the deal will be booked and how the risk group will generate valuation and issue P&L reports on the exposure. The risk group is beholden to approved modeling methodologies and the trading system of record (and whatever engine it uses), in contrast to the more intellectually unfettered commercial groups. Traders may feel that an option is too cheap based on gut instinct, and quant/structuring may agree based on a cutting-edge model or something of its own creation, but if risk is only approved to use Black-Scholes or a binomial tree, neither of which is designed for the product in question, there will be severe internal valuation problems.

Footnotes

[72] Some maintenance decisions in certain industries are definitely driven by economics, which can make the calculus even more challenging.

[73] There are a lot of different option models. Once the trader moves past the well-understood general solutions like Black-Scholes, binomial models, or Monte Carlo simulations things tend to get very product and circumstance specific.

All rights reserved. No part of this book may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by reviewers, who may quote brief passages in a review.