trending_up

Adaptive Investment Algorithms

Data-driven investment, risk management, and beyond

FinTech Legal Services

Expert witness and litigation support in Financial Technology
visibility

Strategic Critical-Path Consulting

Roadmap analysis, project valuation, and due diligence
  News and Publications
Walter gave a tutorial overview of Machine Learning for Quants at the Fall 2017 Q-Group Seminar in Vancouver. Our view is that halfway between under-informed elevator pitches and months-long study of the field, many practitioners could benefit from learning why things are the way they are today, how it relates to what they know, and the decision criteria for putting in some significant effort to build the requisite skill set.
An Intro to Machine Learning For Quant Investment: "Eventually Probably Approximately Correct", Presented at Q-Group Fall 2017 Seminar from Walter Tackett
trending_up

Adaptive Investment Algorithms

Data-driven investment, risk management, and beyond

The assets of the average investor are managed with statistical methods that pre-date the Internet and use simplifying assumptions designed to accommodate mechanical calculators [1]. Aggregating weak or incomplete assumptions into increasingly complex financial models - and treating their deviations from actual data as “noise” – is an historically common practice. In the past thirty years, the mismatch of model and data has contributed to every financial crisis from 1987's Black Monday through 2008, as well as to smaller and more numerous nasty surprises that continue to this day.

During that same period, most famously following the appeal of [Efron & Tibshirani, 1991], scientists and engineers at large have shifted to statistical algorithms that use empirical data together with intensive computation as a substitute for simplifying assumptions and closed-form equations. Whether called Machine Learning, Computational Statistics, or Adaptive Processing, only recently has mainstream investment practice come to recognize the value of combining disciplined non-parametric statistics with high performance computing.

Most firms currently working on Machine Learning and other compute-intensive investment activities consist of specialists from one field scrambling to learn the other: nE12 (pronounced n-Trillion) is founded on our long-standing belief that algorithms and computer science are integral to financial decision making. They hold importance equal to that of economics, accounting, and statistics. Algorithms as a unique category of return and risk are central to our views on investment.

Moreover, investment decisions are only a small portion of the larger transformation that algorithms are rapidly bringing to finance: as digital currencies form new asset classes, the integration of Finance and Computer Science becomes ever more imperative.

[1] Efron, Bradley, and Robert Tibshirani. "Statistical data analysis in the computer age." Science (1991): 390-395.
(Click to copy a link to the paper)
visibility

Strategic Critical-Path Consulting

Roadmap analysis, project valuation, and due diligence

The average investor is more likely to reap the benefits of Machine Learning and Data Science when shopping for consumer goods, or selecting a movie and restaurant for the evening, than they are in the investment of their wealth. Their sensitive information stored on Google’s cloud servers is generally more resistant to hacking than the sensitive information stored by their banks, brokerages, and credit card companies. In its annual report on top financial services issues for 2017, PWC's Financial Services Institute [5] identifies AI "catch-up" as the #1 challenge industry-wide [6], followed closely by other fintech issues, with Brexit (at the time of writing) being virtually the only threat of non-technical origin.

Firms face pressure to field new capabilities, with daunting questions of buy-vs-build, acquisition, and licensing; They are motivated to take on new projects, but unsure of the pitfalls, facing real uncertainty about which initiatives bring expected benefits, and which harbor hidden complexities. Most Financial Firms have not spent the past fifteen years treating Computer Science as a direct source of competitive advantage the way that Google, Amazon, or Facebook do. For most investment firms, traditionally driven by infrastructure, scale, and compliance, the core talent to drive strategic innovation in algorithms and adaptive systems is not in-house. Yet the competitive edge of the “haves” over the “have nots” will be critical in determining tomorrow's winners and losers.

As investment firms race to build competency in advanced computing and hire key talent to make it a sustainable reality, nE12 is a partner who knows the landscape and understands the issues. Our firm is ideally positioned to help build an achievable roadmap and coach its execution to a measurable payoff.

[5] PWC Financial Services Institute, web site, Dec 2016, last reviewed June 30, 2017.
(Click to copy a link to the PWC's FSI website)
[6] Rao, Saxena, et al. (1996),“Top financial services issues of 2017: Thriving in uncertain times,” PWC Financial Services Institute Report, December, 2016.
(Click to copy a link for downloading a pdf of the PWC/FSI report)

The Firm

Please direct inquiries to:

Thank you for your patience as we continue to build out our site and documentation.

Contact : Humans can read this and will know what to do, garden-variety spambots can't.
Copyright © 2017 NXE12 LLC
All rights reserved.