The discipline of financial engineering has been criticized on the ground that it blindly follows statistical and mathematical paths that lead, at times, to the edge of or over economic cliffs. Despite the risks involved in reliance on quantitative correlations and measures that struggle to capture the unpredictability and irrationality that often drives markets, financial engineering has tremendous potential, in private as well as public markets, to close financing gaps that exist due to standard financial product mismatches. This article provides an overview of financial engineering, discusses some criticisms that have been leveled against it and outlines its great potential to reduce market inefficiency in private market financing transactions.
Financial Engineering
No great forecasting ability was required to predict that increases in data availability and computing power would, sooner or later, make inroads into the world of finance. The combination of an exponential jump in available raw data and the means to rapidly collect and sift through it caused a fundamental reevaluation of the relationship between technology and markets and led to the discipline of financial engineering.
Financial engineering, in simplest terms, is an effort to combine scientific rigor with a range of quantitative tools to solve financial problems or to take advantage of financial opportunities. This can involve converting market information into investment positions, asset and derivatives pricing, restructuring cash flows, risk sharing and creating new financial products. These tools can include mathematics, statistics, software and Artificial Intelligence to strengthen financial decision-making ability and accelerate financial decision-making speed.
At first glance, the use of financial engineering to frame and solve financial problems makes a lot of sense. The ability to access and analyze information is at the heart of many financial issues, such as valuation, and the ability to get data, quickly process it and convert it into investment decisions no doubt has the ability to strengthen financial strategy and decision-making. Moreover, the entire investment process is plagued with many types of decision-making bias, psychological reactions and out-and-out analytical mistakes and setting up quantitative guardrails to reduce decision-making errors would seem, at least intuitively, to hold out the promise for better quality decisions over the long run.
Criticisms of Financial Engineering
Despite the many potential advantages of financial engineering, the financial engineering discipline has not been without its detractors. A number of these criticisms have come on the heels of financial disasters that were rightly or wrongly attributed, to some degree or another, on overreliance on and overexposure to mathematical models that were based on statistical probabilities that were drawn from past events. Many disasters are the unfortunate occurrence of what, according to the dictates of history, should not have occurred.
Many disasters are the unfortunate occurrence of what, according to the dictates of history, should not have occurred.
Some criticisms that have been leveled at financial engineering are set forth below.
Data Quality. One criticism of financial engineering, which applies to many forms of financial analysis, is with respect to the quality of data used to generate financial forecast outputs. While it is possible to access increasingly large amounts of data, the access to larger amounts of data does not necessarily mean that the data is better. It is not a logical impossibility that the systematic removal of filters on data and data distribution has caused data quality on the whole to decline rather than improve. Moreover, the accumulation of data can lead to overconfidence which on many occasions is the lull before the storm of financial losses.
Data Inference. A second criticism leveled at financial engineering is that standard statistical techniques which are used for data inference may not only sharply skew where data set means and outliers lie but impose a semblance of statistical stability on data and data relationships that does not reflect the reality of the financial or economic system the data is trying to portray. In other words, not only is the average of wrong answers not a right answer, but it also may have nothing to do with the right question.
Historical Inferences. A third criticism of financial engineering is that a significant amount of forecasting parameters are based on probability distributions that are drawn from past events. One classic example of this is the Black-Scholes options pricing formula which assumes that future volatility, a key factor in the price of a financial option, is a function of historical volatility. However, the fact that an event occurred very frequently in the past does not mean that it will occur frequently in the future.
Further, for practical purposes financial events are not only defined by their occurrence but also their impact and impact is a function of a constantly changing financial, economic and social backdrop. This means that an event which has a probability of (a) and impact of (b) in time period 1 can have a probability of (c) and an impact of (d) in time period 2. The farther the distance between the present and the forecasted event, the larger, less stable and less certain the probability and impact matrix can grow.
Economic Actor Irrationality. A fourth criticism of financial engineering and in particular the superimposition of mathematical models on financial markets is that it does not take into account the massive impact of “rationality ranges” and highly inconsistent psychological responses to financial and economic events. Economic actors can react to different factors, react to similar facts in different ways and make decisions that lead to very different financial action outcomes. These responses, and the different types of momentum that they can generate, can cause markets to move in ways that are essentially impossible for mathematical models to reliably predict.
Nature of the Future. Yet another criticism of financial engineering is the reality that probability distributions assign probability to events that are known or can be inferred from what is known. However, historically the future has not been a simple linear or non-linear change to the set of known variables but rather than introduction of new variables that have no real historical reference points. The future, in other words, is not bound by the sum of the past. Past and future non-continuity is a major blind spot for predictive models that are heavily based on past events.
Market Gaps and Financial Engineering
Despite the criticisms it has faced, financial engineering nevertheless has the ability to significantly improve market efficiency. One way it can do this is not only by predicting market outcomes but by closing market gaps.
A financial market is a set of actual or potential financial transactions. These transactions represent different degrees of economic efficiency as set forth in the below market transaction matrix. Transactions that close or do not close on sub-optimal economic terms represent market gaps. Each market gap creates economic costs that must be absorbed by society in one way or another.
Market Transaction Matrix
| Transaction | Economic Efficiency |
| Closes | Represents optimal result for all transaction parties |
| Closes | Does not represent optimal result for all transaction parties |
| Does Not Close | Represents optimal result for all transaction parties |
| Does Not Close | Does not represent optimal result for all transaction parties |
There are many reasons for market gaps, including the inability to identify deal counterparties and lack of information that is necessary to price transactions. Let us consider how financial engineering can be used to address three reasons for market gaps: (i) duration mismatches; (ii) deal requirement and deal structure mismatches; and (iii) risk and return mismatches.
Duration Mismatches. One way that financial engineering can be used is to eliminate duration mismatches. Companies and investors have different cash need and investment horizons and when these do not match many deals do not close. For example, a company may need funds in May but an investor can only provide funds in June. Financial engineering can be used to remove this gap by, for example, obtaining a bridge loan from a third party who provide funds in May but will be repaid by another investor in June.
Deal Mismatches. Another way that financial engineering can be used is to retrofit financial instruments to fit market and transaction requirements. For example, assume that a company wants low short-term financing costs that would ordinarily be provided by a loan but does not have collateral. An investor has flexibility regarding short-term financial costs but wants the possibility of greater return than provided in a standard loan agreement. Financial engineering can be used to create a new product, such as soft loan, where in exchange for a low interest rate the investor receives a percentage of cash flows up to the point that an agreed rate of return is met.
Risk Reallocation. A third use of financial engineering is to fractionalize risk. For example, an investor may not wish to invest in a company whose cash flows pass through cycles that are considered to be too risky. However, through financial engineering a single set of cash flows may be broken into multiple financial products with different risk and return profiles. This can increase capital options for a company and bring down capital costs.
Conclusion
While financial engineering, as with other disciplines that are used in connection with forecasting, faces significant challenges the core idea has great potential. Apart from forecasting applications, financial engineering can be used to help eliminate market inefficiencies, close funding gaps and significantly bring down the cost of capital. Increasing transaction efficiency reduces the risk of externalities that have negative implications for companies, financial systems and economies.