Are there ways to time your investment?
The quest of trying to time the market never stops. After consuming tons of content and reading articles that tell you “do not time the market”, it is “time in the market” that matters the most. As a passive investor, you would think I would listen to this, but I always have ideas on how I could beat the market by timing my investments. Generally, I ignore these ideas and carry on with monthly contributions to my investment portfolio. However, I decided to test one of these ideas to see if it had any merit.
In the past few years, we have seen the number of retail investors participating in the stock market increase. More online forums and personalities are making investments in the stock market accessible through education. We have also seen an increase in cheap, easily-accessible, low-cost digital trading platforms. These digital platforms have increased the amount of retail investors trading in the stock market.
When Covid hit in 2020, the activity in the stock market increased, and we saw sharp rises in the number of stocks traded per day. For example, the Vanguard S&P 500 (VUSA ticker) has seen an increase in the average volume of shares traded per day between 2020 to 2022 relative to the years before 2020. This is when I had a light bulb moment on a trading strategy that could work.
Generally, if a stock has a high trade volume, the demand from investors could increase the stock price. Some investors stock pick, whilst some invest in Exchange Traded Funds (ETFs), and others do a mix of both.
Some retail investors work a typical 9–5 job, and investing is a way for them to increase their income passively. So they tend to invest in Robo Advisors or choose ETFs that their favourite finance YouTuber or Podcaster has talked about in the past.
So I thought to myself if people invest when they get paid or a few days later, and a lot of these retail investors buy ETFs, this could mean that the trade volume of those ETFs popular to these retail investors might go up on payday or a few days after.
To test this idea I needed to know a few things:
- What are the popular ETFs?
- When do people get paid?
What are the popular ETFs?
I came across an article by Barclays which showed the most popular ETFs which is based on the number of purchases and sales deals placed by our Smart Investor customers in a given week. This list is updated every week so at the time of writing the most popular ETFs were:
VUSA, ISF and VUKE are a few of the most popular ETFs according to the Barclays investor platforms. It is no surprise that the ETFs that track the S&P 500 or the FTSE 100 are the most popular in the UK.
When do people get paid?
This turned out to be trickier than I thought. Some people get paid weekly, fortnightly or monthly. Data from the Chartered Institute for payroll and pension (CIPP) shows that the monthly payment is the most common pay frequency.
Based on my theory, these paydays (+/- 2 days) should have a higher trade volume than non-paydays. The overlap between the last day of the month, the 25th and the 28th made me group them as one period in this analysis. For the sake of this analysis, I have three crucial periods every month:
Strategy A involves exclusively making monthly contributions in Period A, Strategy B involves exclusively making monthly contributions in Period B, and Strategy C involves exclusively making monthly contributions in Period C.
The 25th or the 15th could fall on a weekend or a non-market day. For such cases, the algorithm would choose the last working day before that weekend as the start of both periods.
How did I evaluate the performance of each strategy?
I ran several simulations considering different investment horizons and start dates. Each iteration of the simulation applies all three strategies.
The investment horizon I looked at includes 1–5 year periods. I also considered months from the start of January 2015 to the end of January 2022.
Does trade volume differ across these three periods?
Before running this simulation, I think it is important to see if the trade volumes are different between each of the periods we have described. We look at each year and compute the probability of observing months where the trade volume is higher than the monthly average traded volume.
In recent years, with the rise of retail investors, the trade volume on non-payday dates (i.e. Period C) tends to be much lower, as expected. Although this pattern is somewhat weak pre-2020. This suggests that there the influence of paydays on trade volume is not consistent and our theory has a weak chance of holding water due to other important factors that sway the market. Nevertheless, we still run our simulations.
Do we see any signal from the strategy?
To test this hypothesis I choose two instruments. I ran the scenario model on Vanguard S&P 500 UCITS ETF (VUSA) and I also ran it on Bitcoin. I think it would have been interesting to run this on several stocks and compare all the results but unfortunately, I did not have the time to do so.
For VUSA, Strategy C i.e. investing outside the popular payday periods (Period C) tends to show an additional rate of return of around 0.1%. However, there is not enough evidence to firmly suggest that it is better than Strategy B. On the other hand, investing in Period A (which is generally the last week of the month and the first few days of the month) seems to be the worst period to consistently invest in with a 33% chance of outperforming someone who has no consistent period of investing.
This is probably the result the “crypto bros” were waiting for. Surprisingly, it matters most for crypto. According to the results, never think of buying crypto during Period C (end and beginning of the month). Investing in periods B and C shows that users could get additional returns of 9% and 5% respectively. There is also a high chance of making additional returns.
Trying to time the market by anticipating the trade volume changes that could be driven by payday periods could be a strategy for making additional positive returns, especially for cryptocurrencies, however, this seems to be a less influential strategy for the S&P 500.
