Strategy test – Piotroski F SCORE. How much the guru will earn on LSE?
All investors have heard about a lot of ready to use strategies, which help to select superb stocks. During some periods they work, but very often they don’t. There is not such a thing like “magical formula”, which works most of the time and produces better, than average profit. On the over hand, some ideas seem very coherently and even the most experienced investors cannot discredit, fundamental thinking behind them.
In this article, we would like to test an F-Score indicator and strategy created by an academic J. Piotroski. A lot of similar experiments have been conducted before and we will compare our results with them.
In the beginning, let’s look briefly on this strategy. If we want to use it, we choose the undervalued companies (usually we start with 20% of the cheapest stocks according to P/BV). Piotroski has pointed out, that such companies outperform the market in the long run (20 years perspective), but there are a lot of periods, where they do not. In this situation, he has developed several criteria, which are summed up in one number:
- 1. Positive return on assets in the current year (1 point)
- 2. Positive operating cash flow in the current year (1 point)
- 3. Higher return on assets (ROA) in the current period compared to the ROA in the previous year (1 point)
- 4. Cash flow from operations are greater than ROA (1 point)
Leverage, Liquidity and Source of Funds
- 5. Lower ratio of long-term debt to in the current period compared value in the previous year (1 point)
- 6. Higher current ratio this year compared to the previous year (1 point)
- 7. No new shares were issued in the last year (1 point)
- 8. A higher gross margin compared to the previous year (1 point)
- 9. A higher asset turnover ratio compared to the previous year (1 point)
The biggest score is 9. The higher the score, the better are the company’s fundamentals. Piotroski has proposed a strategy in which we choose the cheapest stocks with a result 8 or 9, because they should be the best market performers in the nearest future.
Firstly, we will check the results of the test conducted on the Old School Value blog (Piotroski Score Screen Performance). You can see, that this strategy has outperformed the S&P index, but do not be excited too soon. If you look more closely on these charts, you will notice, that all the advantage over market was created during the 2003-2007 period (long bull market). The rest of the time, the portfolio based on the F-Score and P/BV value was behaving not better or even worse than the market.
The next test worth to look at, has been founded on the Stockopedia.com (F-Score). Their chart suggests, that analyzed strategy produces some annual return on the market in UK and UE, but we can tell you, that there are not better, than an un-weighted index of all shares (10,45% annualised return, once again we will use this base to compare results of our folios later in this article). It is true, that the examined period is too short to say anything conclusive (especially in the case of the USA; negative annualised return). With high level of caution, we can ascertain, that this strategy will not achieve superb results in the period of 3 years.
The last test we would like to talk about is the most curious one (Piotroski Score). It examines long period of time (1985-2011), so we can rely on the results. The key findings are that, that the higher the score, the higher annual return, moreover companies with score from 7-9 will possible beat the market in the long run.
Further, the author states that, stocks with 9 Piotroski score, return an average of 15% in the US and UK, 18% in Europe irrespective of the price you pay. We need to know more about the methodology of this test, to pertain to it. Especially, with the statement that buying a portfolio of stocks with High Piotroski scores and Low Price/Book Values will have outperformed the index by in excess of 5% per annum (1986-2011). Even though, this results are quite good and favour tested strategy.
Eventually, let’s see our results.
- we included only stocks at price higher than 20 p.
- volume over 8000 thousands p.
- if there is no information about position size, all stocks fulfilling the criteria have been included
Combining low P/BV with F-score, should give us better return over market in the shorter period of time. In the beginning, let’s check how has performed the folio comprising 20% cheapest shares:
It has beaten on average the un-weighted index about 0,25 percentage point annually. In the next step, we will compare results above with Piotroski strategy. We have tested several folios and below is presented the one, with the highest annual return:
The un-weighted index of all stocks makes on average 10,45% during 12 months period, so the best folio has not beaten neither the P/BV folio, nor the market. The question is why? The problem with Piotroski strategy is that, it leaves only a few stocks to trade. There was a lot of periods, when the folios have not been even close to full. Not invested money cannot generate any return.
On the other hand, we have encountered another problem. In few cases, when we have established variable level of our position Piotrosky strategy has beaten the market. Why we excluded these folios, as proof, that it is working? The answer is the size of the position. In these folios, we have had periods when engagement in ONE stock could be even 50%. In this situation, the whole result can be generated by one stock. When we miss to pick up that stock or the right moment to buy, the whole strategy will be useless. We are seeking for a reliable method of creating superb folios and it is for sure, not one of them.
In the table below are presented results of other tests:
All folios have been tested for the same period of time.
- Piotrosky strategy does not provide higher earnings,
- F-score is useful in pointing out very good stock from the whole market. Folio consisted with them, can generate better than average returns (folios 7-15),
- F-score can be used for creating a defensive folio, which will have lower maximal drawdown, than the market. Stocks with a score 8 and 9,
In case of using Piotrosky strategy for creating the folio, our findings are the most convergent with the tests, that discredit effectiveness of it. However, we can affirm, that F-score is very handy tool in pointing out superb companies, but not from the group of the cheapest ones according to P/BV indicator. Strategy based solely on the F-score between 8 and 9, or even 7-9 can provide better, than average returns in the long run and be a shield against harsh and sudden bear markets.
Below you can find companies, which are currently pointed out, by P/BV (cheapest 40%) and F-score (7-9). As you can see, even with so extended range we have 33 stocks to trade; only one “9” and seven “8”.
If we apply Piotroski strategy in the original version and loose our criteria about volume and minimal price, we will find ZERO buying signals. These criteria are very severe and even on such wide market as LSE, you will not be able to find any matching stock. As we have said earlier, F-score is great indicator to look for healthy, well-performing companies among all tradable stocks, not only the cheapest one.
Other articles worth to look at: