Industrial Revolution (2002)

18 Celebration of Excellence: Sponsor’s Foreword

April 2002

An industrial revolution: new capital markets paradigm

B

ankers have the unfortunate habit of thinking of themselves as the best and the brightest — products of unemotional meritocracy — creative, entrepreneurial, calculated risk-takers. After all, wasn’t Sherman McCoy (Tom Wolfe’s apocryphal bond salesman) a certified master of the universe? Although his world came crashing down around him for reasons not directly related to his job performance, his real world peers are nonetheless fortunate that their jobs require something far short of perfection in execution. In a recent speech, Jack Welch, the former chairman of General Electric, made exactly this point: “...[if] you put six sigma in an investment bank, they would all gag!” In case you think he was just engaging in some gratuitous banker bashing, consider this: six sigma quality means having fewer than 3.4 defects or errors per million operations in a service process. That is 99.99966% perfection. Contrast this benchmark with the assurance once made to me — by a senior syndicate manager of one of the largest and most respected global bond underwriters — that it was perfectly normal and necessary to expect and reserve for 5%-10% errors in the allocation of a jumbo multi-tranche bond deal! Assuming an average of 200 individual orders (including splits) on a typical new issue, to reach six sigma quality levels you would need to have fewer than four errors over 5000 issues! Some banks are lucky if they have less than four errors per deal. Even if you change the assumptions, it is obvious that our industry is embarrassingly tolerant of poor process execution. So why do customers and clients accept this level of service? Firstly, the direct costs are largely borne by the banks themselves, and secondly, it is very difficult for a customer to objectively and quantitatively measure the quality of service. (Compare the problem of an investor or issuer in identifying variation from best/perfect execution on a new capital markets issue to that of a computer manufacturer, for instance, who is supplied with a shipment of 100,000 microprocessors of which x are defective.) Assuming that the banks cannot and do not indirectly pass on these costs, why have they not made more significant progress in addressing this problem?

A plea for perfectionism My guess? In one word, hubris — fixing this is all about the details. It requires painstaking attention to process, to the banal minutia of the obvious and an unswerving dedication to perfection —

EuroWeek

qualities more often found in engineers than masters of the universe. Fixing these shortcomings is just not seen as exciting or important to teams of bankers feverishly searching for the next million dollar trade or idea. However, intuition tells us that even if the costs are difficult to measure and are absorbed by the banks, they are no less real for this, and ultimately result in one or all of the following: higher capital costs for issuers, lower investment returns for investors, and lower operating profits for banks. And therein lies the next major opportunity for capital markets bankers over the next decade: to use technology not only as an enabler of innovation (as has been the case over the past 15 years) but as a driver of industrial efficiencies.

Mapping out a paradigm shift So what exactly does an industrial revolution in investment banking (or more accurately defined, capital markets) mean? Why is this relevant now? What are the advantages for issuers, investors and banks? What are the costs? Who will be the winners (and losers)? Is technology and/or the internet an enabler of the revolution? And if so why?

Our industry is embarrassingly tolerant of poor process execution The industrial revolution in investment banking is all about creating a new paradigm for the execution of capital markets business. It is about reinventing the organisational mindset, replacing the traditional front, middle and back office with a highly flexible and efficient product factory attached to a professional cadre of relationship managers and solution providers who work with customers and clients to tailor products and solutions to be produced and executed by the factory. It is about viewing the services we provide as two distinct value propositions, one resting on the creativity and knowledge base of the bank and its bankers, and the other resting on the efficiency and accuracy of production and execution. Two factors are driving the inevitability of the paradigm shift: market size and costeffective technological solutions. The incredible growth and globalisation enjoyed by capital markets over the last 15 years has given us a market where economies of scale

Sean Park, global head of debt syndicate, Dresdner Kleinwort Wasserstein

start to become extremely compelling: rather than dealing with tens of issues, hundreds of tickets, and tens or hundreds of millions of trading and underwriting, the typical medium to large sized investment bank now deals with numbers that are a factor of ten or more higher than these. The upfront and fixed costs of building the (highly scaleable) factory are now justified by the volumes of throughput. Hand in hand with this growth in volumes, the market has seen increasing standardisation including, but not restricted to the introduction of the euro, debt issuance programmes, benchmarking and so on and so forth. As a result, for even the most bespoke products and client solutions, upwards of 80% (often more than 95%) of the process can be stripped away and completely engineered to very high tolerances. At the same time as the scale and homogeneity of the market opportunity have increased, the (technological) inputs needed to build the factory have become more and more useful, flexible and inexpensive, reducing not only the costs of building the factory in the first place, but vastly improving time to initial delivery and limiting future maintenance and capex requirements. The coming together of software, novel programming techniques, processing power and communications infrastructure witnessed by the birth of the internet (or any other private network for that matter) is and will continue to be the technological catalyst for an industrial revolution in capital markets. One of the reasons why banks have been

www.euroweek.com

April 2002

slow to embrace (wholesale) process reengineering in their capital markets activities is that the incremental improvements (in costs, revenues, business velocity, capacity utilisation, etc) are in many cases relatively small — sometimes invisible — when considered in isolation. Fixing one part of the process chain in isolation may save you half a cent in the dollar. Fixing the whole process chain, however, allows you to cut real costs and raise productivity, by reducing headcount or reallocating human resource to higher value added customer facing activities. The key to success lies ultimately in taking a holistic approach, with real benefits accruing geometrically as the weak links in the value chain are eliminated. Consider as an example the new issue process. I take this as an example as it is probably one of the most obvious, most straightforward and best understood of the hundreds of processes executed in any investment bank: pitch, mandate, documentation, marketing, bookbuilding, allocation, pricing, trading. Any first year associate could write the detailed process flowchart. The possible variations (currency, documentation type, issuer type, etc) are ultimately all more or less elaborate forms of the same base case. (The underlying process used in executing a 144A deal in the US for a high yield company is hardly different to the process used to issue an FRN for a bank off of an EMTN programme.) So why is it that this is still such a labour intensive, error prone and ad hoc process across the industry?

Tools of revolution Over the past 12-18 months, the seeds of change have been planted to address this situation. Tools such as Dresdner Kleinwort Wasserstein’s e-bookbuilding have pioneered a more professional and industrialised approach to the new issue execution process. One of the most important innovations was to open up our system to allow all participants in a new issue to communicate on a common basis. This was a necessary but far from sufficient first step.

Over the past two years a whole industry has emerged to build bridges between market participants Perhaps the most valuable longer term contribution of e-bookbuilding has been the ability to clearly identify and understand the details and bottlenecks in the new issues process, which are often less than intuitive and have emerged by using the system in live production now for almost two years. As indicated above, the basic new issues workflow or framework is relatively straightforward and easily defined. The devil, however, is in the details. An example of one of the bottlenecks discovered through the use of e-bookbuilding

www.euroweek.com

Celebration of Excellence: Sponsor’s Foreword 19 Banks have Formula 1 drivers racing around in anything from a Trabant (for the worst) to a BMW 3 series (for the best), without any pit crews and other similar utilities is the lack of common investor identifiers. Just as securities have common codes (Cusips, ISIN codes) it is important in the context of efficiently measuring demand and allocating new issue securities that investors be easily and uniquely identifiable in real time (imagine the chaos that would ensue if every market participant had their own unique identifier for each security). The need for a common, unique investor identifier has now been accepted by the market as self-evident; as late as six to nine months ago it was hardly on the radar screen! Indeed, underwriters are now collectively acting to address this omission. A second issue that has now been made obvious is the necessity for the various banks’ new issue utilities to be able to communicate electronically with one another (and ultimately with other participants: issuers, investors, press, stock exchanges, etc). While e-bookbuilding initially solved this problem by opening its platform to competitors and clients alike, this was and remains a sub-optimal solution. Given that most banks have their own new issue toolsets (even if for the least advanced this can consist of as little as an Excel spreadsheet) the duplication of effort (and concomitant scope for error) needed for data entry should reasonably be avoided. What is needed is the ability for systems to communicate in real time. This can be solved by creating individualised bilateral APIs between systems, or more rationally by agreeing a market standard common new issue syntax that when used, for instance, in an XML framework would allow all systems to communicate effortlessly, irrespective of their individual functionality. Indeed DrKW has recently circulated a draft syntax to a number of other major investment banks for comment and discussion — due to our extensive electronic deal experience we were able to produce this in a day. But don’t overestimate our contribution or generosity: in essence, it is simply a list and standardized formatting for the approximately 50100 variables needed to accurately define a new issue and the bookbuilding and allocation process (of which some of the more obvious are: issuer, coupon, currency, maturity, order size, order status, split amounts and so on). Where technology is important, many — perhaps most — of the big investment banks have much of what is needed, or could easily build or buy what they lack. What is most often missing are the tools to put the pieces together, whether that be internally, externally or both. Ironically, as evidenced by the example above, the cost or technical difficulty of building these

tools is often relatively trivial. Indeed, seeing this, over the past two years a whole (rapidly consolidating) industry has emerged to build bridges between market participants, with business models often consisting of no more than standardizing trading and data communication for connecting multiple counterparties. As communication protocols, syntax and data formats become standardised in the public domain, many of these companies will find their transaction-based value proposition disappearing (if it ever existed) and survival, if even possible, subject to switching to a software-licensing business model (analogous to Red Hat for Linux) or a fully-fledged exchange model. (Moral: beware of people selling you what is essentially available for free.)

Changing a bank’s culture So if technology is not a fundamental bottleneck or differentiator, what is? Culture and attitude. It comes back to attention to detail and a determination to impose process standardisation and to demand quality execution. The potential rewards are strikingly obvious: higher productivity, lower fixed costs, higher customer satisfaction. But the rewards will only accrue to those banks who embrace this vision with an all or nothing mentality. Anything less will result in a situation all too familiar to many investment banks: higher costs and lower productivity, as the benefits of better process engineering, being incomplete, are impossible to harness. Success depends on unconditional institutional adoption of this new paradigm. I have taken only one very small example of a standard capital markets process. There are hundreds more repeated thousands of times daily, some more efficiently than others, but none I would guess coming anywhere close to six sigma quality and efficiency. As valuable as re-engineering an individual process may be, the real opportunity comes from the prospect of using this new paradigm and technology to weave them into an ever more powerful and efficient whole greater than the sum of its parts. The challenge ahead lies in building the best capital market factory in the world. Nice theory, but you’ll never get bankers to think of their business in this way — so let’s think about it another way. To use an automotive racing analogy, banks have Formula 1 drivers racing around in anything from a Trabant (for the worst) to a BMW 3 series (for the best), without any pit crews. Don’t you think the first bank to get an Aston Martin Vanquish or, better yet, a Formula 1 car might win — even without a Schumacher in the cockpit. Now what if I told you that with the right connections and a little determination, the Vanquish could be had for the same price as the 3 series...? Good bankers are just too expensive to waste driving the equivalent of a Trabant. Investment banks don’t have to embrace the revolution, it’s just that they can’t afford not to. ❙

EuroWeek