08 26 2019 | by Victor Xing | Economics
07 12 2019 | by Victor Xing | Capital Markets
Kekselias portfolio one-year return: 51.5%
02 27 2019 | by Victor Xing | Economics
Common catalyst for progressive and conservative populism
12 09 2018 | by Victor Xing | Capital Markets
Kekselias performance review: 1.31% YTD total return
10 14 2018 | by Victor Xing | Capital Markets
Roundabout path in the snap-back of long-term bond yields
09 23 2018 | by Victor Xing | Central Banks
Calm before the storm as quantitative tightening looms
05 20 2018 | by Victor Xing | Central Banks
Alternative narrative on the natural rate of interest
01 07 2018 | by Victor Xing | Capital Markets
Flatter yield curve a symptom of ineffective tightening
12 04 2017 | by Victor Xing | Central Banks
Bond market term premium and wolves of Yellowstone
10 17 2017 | by Victor Xing | Capital Markets
How we learned to stop worrying and love the “fake markets”
11 13 2015 | by Victor Xing | Financial Models
Is financial engineering useful?
Financial engineering is a really broad term that covers a diverse field of responsibilities. Some of these responsibilities are very important (and useful). It is a great experience for anyone who can remain focused for hours at a time. Rather than defining what “financing engineering” is, I would like to use my past experiences as an example. It is worth-noting that few people called me “financial engineer” – it was often “analyst” (the most junior title on Wall Street) – and my favorite one was “junior boy.” I always fetched coffee orders correctly for Boss Man.
When I first arrived at the trading desk, I was a tech-support person with a programmer background. Many people thought it was a terrible idea for me to leave my programming job to “hook up monitors, replace mouse, and ask traders to reboot their machines.” However, I really wanted to be the fly-on-the-wall to learn more about the core business of the firm, which is not trading application development (my former role). I was willing to do anything to get that experience.
Financial engineering – model upgrade and automation
From there on, I worked on a number of Excel and MATLAB models for traders and portfolio managers, and it became more complicated as time go by. Evolution of a model upgrade often take the following path:
- [Existing Excel model] – someone would manually load data into Excel and change formulas daily to derive model results, then screen grab to send out results to PMs via email
- Phase 1 objective: the person only needs to open the spreadsheet to complete all work that used to be manual. automate data retrieval via Bloomberg or other financial data providers. Change hard-coded formulas to dynamic, self-adjusting implementations. Clean up look and feel. This frees junior people from the mind-numbing task of updating spreadsheets.
- Phase 2 objective: remove the need for a human operator to open the spreadsheet. Use MATLAB to open Excel in the background, load spreadsheet, trigger data update, scan for result value, capture value, save and close Excel via API, store value (or array of values) into MATLAB matrix. Use build-in charting functions to create daily rolling charts. Initiate 3rd party email program in the background, dump the latest chart into template, trigger send. Voilà – portfolio managers get latest results without needing human interaction.
Investment decision engines
Other times, the model would capture the essence of a PM’s investment thesis (these are really fun to work with). The model would take in all the time series and market feeds the PM would ordinarily look at, and they will be sent to a central “decision logic” to determine whether an asset’s market price is rich (or cheap) vs. valuation based on the PM’s assumptions. Such models are highly proprietary (some may remain unknown to other PMs for an extended time), and I learned more about investing by working on models than in classes.
More about– it is charting function is versatile. Below is a chart created by my former colleague who now works at the New York Fed – this is all done via MATLAB.
The work is very useful, rewarding, but it is by no means slow-paced. It is the combination of computer programming (designing of an automated Excel spreadsheet often use the same workflow design concepts in programming) with finance. Since investment decisions may be based on these models, downtime must be minimized. Once a design is finalized, it is “work your butt off from 2pm (market close) to midnight” so it will be ready for the next trading day. I remember getting 4 hours of sleep for a week or two just to cram in as much work as I could between trading sessions.
I do find the term “financial engineering” a bit fancier than what the work entails, and the role often overlaps with people with other responsibilities (trader, analyst or junior portfolio managers). For example, later in my career I was helping to manage assets while working on models on the side.
I hope this provides a few examples of whether financial engineering is useful, and I would like to clarify that many people may have vastly different experiences on what the work entails or what financial engineering actually is.
Next article11 13 2015 | by Victor Xing | Capital Markets