| Author||V.T.H. Hoogervorst|
|Title||Inside the law of one price and limit to arbitrage : the case of Heineken and Heineken Holding and others.|
|Faculty||Faculty of Economics and Business|
|Abstract||This paper is a clinical study documenting a systematic violation of the law of one price. The same cash flow are available from of two different stocks, yet the prices of each stock always diverge (between 6% and 40%) for as long as data are available (more then 30 years). Stocks under investigation are large capitalization and they are traded on the same stock-exchange and the mispricing is common information and trivial to compute. We study the determinants of the mispricing, show that arbitrage profits before transaction costs are huge (above 100% per year with zero beta) and assess the arbitrage costs. We study events of particular interest to gain insights into mispricing dynamics such as earning announcements in each stock. Finally, we investigate if mispricing is related to liquidity risk of two different stock.|
This paper reports cases of the violation of the law of one price that lasts for as long as data are available (30 years), concerns large stocks and the violation occurs on the same stock exchange (i.e. is unrelated to the location of trade) and this mispricing is common knowledge.
We find that the deviations from the law of one price are severe and clear. Data snooping concerns are at a lowest because we consider very economically natural pairs of stocks. Deviations are both unrelated to standard systematic risk factors and unlikely to be related to unknown or non-standard systematic risk factors. In short, we believe that we show the clearest example of a systematic violation of the law of one price in stock-markets.
The literature shows compelling examples of implausible stock pricing. Nonetheless, the skeptics typically wonder whether these examples are the rule or the exception. For example, most of the cases are drawn from the high-tech mania period of the late 90s. During this period, firms changing their names to dot-com would see a surprising surge in their stock price, and certain firms had a negative implied market value.
Stocks traded on the Amsterdam stock exchange and Brussel stock exchange sheds light on this issue. When we daily observe two stock prices that have the same value under the market efficiency hypothesis, the two stock prices are different. The law of one price is systematically violated. It seems to be the rule rather than the exception.
It is intriguing to prove directly that the law of one price is violated. This will be difficult because stocks are very similar and their prices are co-integrated in efficient market then there should not be any trivial predictability in returns. This paper precisely shows empirically the existence of such a predictability.
This paper examines seven pairs of Dutch and Belgian stocks and show the expected risk and return of four pairs from a simple pair trading strategy. The return is staggering and the risk is small and idiosyncratic. In some cases, an investor arbitraging from 1973 to 2006 would have multiplied her wealth a hundred thousand times.
We investigate the following pairs: Heineken Holding (HBA) and Heineken (HB), Dordtsche Petroleum (DP) and Royal Dutch (RD), Calve Delft (CD) and Unilever certificates (Uni), Arnhemsche (ARN) and Akzo (AZ), Almancora and KBC Group, Tubize and UCB and Solvac and Solvay. For each pair, the former holds passively a stake in the second stock. As such, HBA, DP, CD, ARN, Almancora, Tubize and Solvac could then be thought of as passively managed closed end funds.
Closed end funds’ problem is that intrinsic value of the fund cannot be known for sure. The advantage of our data is that we can observe the discount at a daily frequency, we now can evaluate the intrinsic value of a stock with precision. There is now sample evidence that stock prices can be wrong. We want to attribute to the fact how much they can be wrong by investigating some explanations like the costs component of a holding structure, that stocks are inconsistent because they co-move with market segments (Froot and Dabora, 1999) and they co-move better when news event occur like earning announcements, information salience and the difference in the expected returns of two stocks may be explained by their liquidity. We can only explain a marginal part of mispricing by costs of keeping a holding structure. We show that large caps in the Netherlands and Belgium are obviously mispriced because they co-move with other segment in the market. Stocks are mispriced because they co-move more with each other during earning announcement periods. We can show that Belgian large caps are much less mispriced than the Dutch large cap and we find no evidence that information salience has influence on the mispricing.
The example of RD and Shell and other shows that prices can be wrong for an extended period of time and by a large amount. These example being large capitalizations, it might be an indication of more mispricings in the market. This paper shows cases similar to RD and Shell except that both stocks are traded on the same exchange, and, interestingly, most cases got arbitraged after many years of mispricing. Some cases still survive and still present anomalies till today.
The remainder of this paper is organized as follows. Section 1 provides information about data, Dutch and Belgian tax laws and an introduction of several stocks who are included in this study. Section 2 describes the statistics of the stocks under investigation and the next section represents the computation of the discounts. Section 3 shows empirical results of the research on possible explanations of the discount and section 4 provides conclusions.
|Document type|| scriptie master|
Use this url to link to this page: http://dare.uva.nl/en/scriptie/215227
Contact us about this recordNotify a colleague
Add to bookbag