Niall Ferguson Quotes

If each of your time steps is one week long, you are not modeling the stock price terribly well over a one-week time period, because you are saying that there are only two possible outcomes.

One of the things we did at PayPal was collaborative filtering and machine learning: looking at patterns of human behavior. We used it there to predict when people would try to cheat the system to get money. But you can predict pretty much any behavior with a certain amount of accuracy.

You can build a statistical model and that's all well and good, but if you're dealing with a new type of financial instrument, for example, or a new type of situation - then the choices you're making are pretty arbitrary in a lot of respects.

Big data has been used by human beings for a long time - just in bricks-and-mortar applications. Insurance and standardized tests are both examples of big data from before the Internet.

Today, I think a CFO needs to be more of an operating CFO: someone who's using the financial data and the data of the company to help drive strategy, the allocation of capital, and the management of risks.

I think the rise of quantitative econometrics and a highly mathematical approach to risk management was the obverse of a decline in interest in financial history.

My career in academic research has not been involved with active management of securities. I've tried to understand risk-and-return relationships; also the pricing of derivative securities.

Insurance firms have always carefully studied real-world data to figure out what, precisely, constitutes a risky activity.

We live in a world where there are many risks, and it's high time we start taking seriously which ones we should be worried about.

One important measurement issue concerns the fat tails problem that I mentioned earlier. VAR is concerned with extreme outcomes. If the tails of the probability distributions we are using are too thin, our VAR measures are likely to be too low.

The global financial crisis - missed by most analysts - shows that most forecasters are poor at pricing in economic/financial risks, let alone geopolitical ones.

Risk comes from not knowing what you're doing.

I had some of the students in my finance class actually do some empirical work on capital structures, to see if we could find any obvious patterns in the data, but we couldn't see any.

Future forecasting is all about testing strategies - it's like a wind tunnel.

One thing I've found is that if you educate and show people the risk, they will do something about it.

When I started out as an equity analyst, we had no securitization data. We relied on company data.

You need to have a lot of human judgment involved in the financial industry in terms of risk management, in terms of investment decisions, and things that really allow us to blend the best of technology and the human brain.

I did long-term re-insurance claims. Asbestos, health hazards, pollution. It was very boring. But I've got quite a mathematical brain and it paid well.

Risk taking in business is one thing. Risk taking in your personal safety is a different thing.

Take calculated risks and limit your downside.

I've said many times that I'm not one of those people who check statistics, follows history, and says, 'If I do this, what will happen.'

The reality is, risk is variable. Those in the financial world know it.

I think the people who've been the most overconfident in our business in the last decade have been the people that called themselves risk managers.

If you don't invest in risk management, it doesn't matter what business you're in, it's a risky business.

Risk is good. Not properly managing your risk is a dangerous leap.