Emerging Quantitative Finance Themes: Machine Learning, Factors, and Rough Volatility

# What are the biggest new advancements in the field of quantitative finance in the last 10 years?


# What are the biggest new advancements in the field of quantitative finance in the last 10 years?












Examples like new research or findings but also older research or findings actually starting to get implemented. Is machine learning starting to get a foothold in finance? Are there any new practical differences in the market, such as new instruments being traded or new regulations having a large impact on the world of finance.

In short, what topics, if any, will be in finance text books 10 years from now that aren't in them now?

Bonus question, how old does a text book or article on finance have to be for you to take that into consideration when picking it out i.e. how old does a book have to be for it to be outdated?

## Answer by Hamish Gibson (score 4)

https://quant.stackexchange.com/a/54014

So to answer your point about ML, I would say yes, it is starting to get a foothold in finance. However, this is in the sense that it starting to be more applied in a particular context. What I mean by this is, it is being used to do old things in a new way...

A particular example of this, is how ML is being used in Risk analysis and Risk Management. The source here by Moody's states that it is powerful and highly accurate. And is becoming more prominent too by credit and quantitative analysts. But the thing is, Risk Analysis and Risk Management has been around since the dawn of credit.

Another particular context is the field of Algorithmic Trading, the source here by JP Morgan notes how ML can be used to yield better returns on investments through using ML to find new and uncorrelated patterns or returns in an asset. In other words, algorithmic trading isn't a new thing, but using ML is a merely new and nuanced application of finding $\alpha$.

I hope that these 2 examples provide an answer as to how ML is being used in the field of Quantitative finance. I am more than happy to discuss in the comments section below.

## Answer by JohnDoe (score 2)

https://quant.stackexchange.com/a/54034

Regarding option pricing, the topic of rough volatility (https://sites.google.com/site/roughvol/home/risks-1) could very well be found in textbooks in ten years. If it is broadly applied in practice is a question I can't answer.

## Answer by Dhruv Mahajan (score 1)

https://quant.stackexchange.com/a/54035

I am going to talk regarding the Asset Management industry here in points.

1) Factor Investing and Strategic Asset Allocation: I am clubbing these two together because asset allocation now is largely influenced by factors on various levels. Factor investing isn’t as new as some fin-media might lead you to believe, Factor research has been in academia for years, but I believe the most significant application to investing was when Cliff Asness started AQR in 1998. But it did take off this decade and I’ll mention that research here. Over the years they have published some amazing white papers which you can read at https://www.aqr.com/Insights/Research . I believe the most defining paper on Factors in this decade would be “Value Momentum Everywhere” by AQR and NYU that cemented the existence of risk premia for these factors across the world and also the diversification effect of value and momentum together. Another amazing paper would be “MSCI Multi-Asset class factor model” , it broadly lays out the implementation and strategy for implementing asset allocation across the globe using various factors like inflation, value etc. across asset classes like Eq, FI, Private Eq , REITs etc. Another mention would be use of Carry factor across asset classes defined in papers like “Carry” by AQR.

2) Tactical Asset Allocation: I am separating the short-term allocation because the research here varies a lot, from use of Factors (BlackRock) to something like Deep Learning (Quantopian). While there is plenty of open research on Strategic Allocation, Tactical remains propriety for most part (You can find white papers for above 2 with a google search though). Unlike the first point I cannot point to “concrete research” in this part since it’s just not cemented enough. You can search for “Global Tactical Asset Allocation” on SSRN and you’ll find plenty of papers, try to implement them and see which works for you.

3) Machine Learning : Instead of me pointing out the plethora of research here, I’d suggest you read what I believe is one of the best books on the topic “Advances in Financial Machine Learning” by Marcos Lopez de Prado. I find the back-testing part in this books one of the most realistic, worth a read.