Disclosure: The article was written on Sunday, February 13, 2011 at 9:18pm. It reflects the writer's personal opinions based on own analysis. The writer is not responsible for decisions taken on the basis of this article. Send your views at kh.asif@gmail.com.
Background
In some countries of the world, issue managers are not allowed to make financial forecasts because by doing so they risk the chance of getting sued (if the real numbers do not match with the forecasts). One of the country that practices this is USA (here are links to two relevant studies on this matter 1. http://findarticles.com/p/articles/mi_m4130/is_3_36/ai_n24222147/ 2. http://ideas.repec.org/a/kap/rqfnac/v26y2006i3p275-299.html ). If I am not mistaken, recently the SEC in Bangladesh has also removed the provision to use forecasts in any prospectus/IM. This means that no type of DCF or Income based approaches can be used for valuation purpose. The only other valuation method possible is thus “relative valuation”.
Trying to use relative valuation methods in frontier markets like Bangladesh lead to some unique challenges. I would like to discuss some of these challenges and maybe try to give some solutions to these that the SEC can take into consideration. The reason I am mentioning SEC is because in our market investor knowledge is limited and other than very few practitioners of valuation most people will be unable to understand what is going on. Currently for the people in the Investment Banking industry there is greater incentive to distort valuation methods (and hence get more deals) to show higher prices for IPO’s.
Challenges and problem
1. Lack of proper comparables: The first step in relative valuation is choosing proper comparable companies. The idea is that similar company should have similar margins and growth prospects and should thus trade at similar multiples. This is a big problem in Bangladesh because the number of stocks in the capital market is small and many companies do not have any comparables at all. Especially for companies with unique project based business model like shipbuilding (Two shipyards are it does not make any sense to compare multiples with companies that have smoother earnings.
2. Presence on non-recurring items in the income statement: This is not really a challenge but rather something that issue managers are doing on purpose to intentionally jack up values. I have observed companies even forecasting capital gains from stock market (present in the historical also) where the company is actually engaged in hotel and entertainment services. Non-recurring items in the income statement would distort earnings multiples. The problem is going to be greater if the comparable companies do not have non-recurring items (leading to higher multiples) while the issuing company has such items (leading to higher issue prices).
3. Book value enhanced by revaluation surpluses: When we observe that new companies have more than 30-40% of their equity in the form of revaluation surpluses the company automatically becomes a suspect. This would be less of a problem if all the companies (included the comparables) revalued assets in the same time frame. However this is not the case and we see that most multinational companies revalue assets after considerable time periods. So naturally these companies will trade at higher P/B’s (and have really high ROE numbers) and using these high P/B ratios will lead to inflation IPO prices for the company that has revaluation surpluses.
4. Earnings cycle and stock market cycle not taken into consideration: This is a major problem of relative valuation in general. A company at the peak of its earnings cycle/business cycle tends to trade at not only its peak earnings multiple but also has its higher earnings. Similarly at the peak of stock market cycle because of the beta effect all stocks tend to trade at high multiples (thus many of these companies are overvalued). So automatically the company is priced very highly and in many instances much above the fair value when the market is doing good and vice versa. This causes a substantial deviation from fair value.
5. Multiples not linked with operating metrics: At the end of the day we buy companies by comparing stock prices with core operating performances. If price multiples like P/E, P/B, EV/EBITDA are being used, they need to be compared alongside operating metrics like sustainable earnings growth (for P/E), core ROE (for P/B) etc. Naturally if Bank A can achieve higher ROE compared to Bank B then we should be ready to pay more for it. Unfortunately, I do not see multiples being justified by profitability and growth prospects.
6. IPO price and opening day stock price seeing considerable deviations: This point is not exactly relevant to relative valuation but I wanted to discuss it because of its importance. As a practitioner of valuation our aim should be to find asset values that are fair to both buyers and sellers. Historically under fixed price methods IPO’s have been significantly underpriced. In recent times with the introduction of direct listing and book building methods we saw the extreme opposite happening. Examples of such direct listings include Jamuna Oil which caused substantial losses to investors. Similar things can be said about KPCL which came through the process of book building. I think a number of things were primary responsible for such gross distortion. Firstly, years of IPO underpricing made investors believe that all IPO’s are profitable. Secondly, investors have a habit of comparing Face Value with issue price which is an absolutely ridiculous idea. Most importantly, there is substantial evidence to conclude that price and volume manipulations are rampantly practiced in Bangladesh. The easiest way to do that is giving out private placements to influential investors who in return keep prices high until the time it takes for retail investors to be convinced that the stock prices will remain high. Unfortunately for the retailers the prices never remains at that level.
Solutions
Now let us look at potential solutions.
1 & 5. I am linking these two together because the solution is similar. Firstly, if we cannot find proper comparables in the domestic market we need to find comparables in other frontier markets (preferable in the same region as Bangladesh). However, it does not end there because despite being in the same industry companies can have different profitability and growth potential due to macroeconomic differences. So, the issue manager can ask for higher than regional multiples if he can justify higher growth and profitability for the company to be listed. Another thing I want to emphasize is the choice of proper multiples. For project based companies like real estate and ship building earnings multiple does not make much sense. Rather a NAV based valuation would be better. For financial companies P/B is the correct choice. For other manufacturing companies P/E, EV/EBITDA, and industry specific (EV/reserves for E&Ps and EV/tonne for cement for example) multiples are appropriate.
2. This is very simple. For all companies, non-recurring items from the income statement has to be removed and multiples should be calculated only on core earnings.
3. Revaluation surplus could be removed from all companies while calculating multiples. The other option is to compare P/B with ROE (Calculated using averaged equity). This way companies with higher revaluation surpluses will automatically be showing lower ROE. In case companies show surpluses they need to provide precise information on how much land bank they have and what is the average price per acre (or any other metric) they have assumed. This will allow investors to see whether the numbers make sense.
4. This is a problem that is very hard to solve. At the peak of the stock market cycle IPO's will always be expensive and at the bottom they will be cheap. The only solution I can suggest is that Issue managers give 3-5 year average P/E, P/B etc of comparables that can be expected to cover a significant part of the business cycle.
6. I think that private placement should actually be banned in a country like Bangladesh where are issues are oversubscribed. In case it is not banned, the lock-in periods have to be increased. Furthermore, the regulators must monitor the trade patterns of private placement holders (the large ones) to identify price and volume manipulation.
Please share article if you have liked it.
No comments:
Post a Comment