Historical data analysis for investment is a big and growing domain. Despite my academic study, I’m still an outsider. I do read such analysis when I consider some stock/bond investments. So I kind of recognize the value. I believe the debunker research more than the marketing propaganda research. Here are my reservations on the value of historical data analysis for personal investment purpose:

Reservation: sample size is small in terms of length. My investment horizon is measured in decades, but how many decades do we have data for? 4 decades? I feel any stock or bond market data more than 40 years ago are not relevant since the environment has changed too much since.

Reservation: Markets are way too correlated. There could be 300,000 stocks and 30 million traded securities (mostly derivatives), but 99.99% of them are not worth analyzing given their correlation to the major market indices.

Reservation: If I have 200k I would plan to invest in properties over a few years, so I dare not put them into stocks. If I only have 100k, I don’t want to spend too much time analyzing.

Reservation: (Evidence) the big mutual fund houses employ armies of highly educated and experienced researchers, but most of their funds can’t beat the U.S. stock indices. How valuable is the research?

Reservation: Picking Mutual fund requires very little data analysis. Based on personal experience investing in mutual funds, I find them easier than stocks.

• Carefree — No need to watch bad news about individual stocks since the fund manager should do that for me. I can just buy and hold and basically forget about it.
• Diversification — Some form of Diversification is in place, but this is a double-edge sword. The high return on individual stocks is often unachievable due to the diversification.
• Lower commitment — I usually start with SGD 1000 and slowly build up. With many stocks I would have to take on exposure.
• Compliance — No issue with Compliance department

In my cynical view, “professional” analysis on historical data for the purpose of picking stocks is a marketing or academic effort.

There’s so much financial data available, so we the laymen feel there must be some hidden values to be extracted. The analyses output do look impressive and valuable. However, due to various factors (include my 1st two reservations), the analyses are not producing better investment returns.

One of my respected investment firm is the Singapore GIC. https://tanbinvest.dreamhosters.com/2015/08/02/gic-2015-update/ is my blog about their performance. If the “market return” over a 20Y horizon is typically 6%, how much “excess return” could I aim to achieve? In other words, what kind of alpha can I get? I doubt my analysis can give me 7%. https://tanbinvest.dreamhosters.com/2017/02/21/6-realistic-long-term-return/ is a blog I wrote on “realistic return rate”