Cross-market trading
There is no need to confine our thinking to just one market. We can envisage a market eco-system that includes not only different types of capital instrument, but also derivatives and markets in different countries.
Strategies such as capital-structure arbitrage, for example, involve trading across more than one market in the securities of a single company. It could, for example, involve buying bonds issued by a company but selling equity in that same company. Although designed to exploit arbitrage anomalies between markets, these strategies can be risky. Yu (2006) examined the expected returns and risks from capital structure arbitrage and found that significant losses occur with alarming frequency. Ofek et al. (2004) have examined violations of ‘put-call parity’, a ‘no-arbitrage’ relationship that one expects to hold in options markets. These violations represent a pricing anomaly between equity and option markets and should be arbitraged away soon after they arise, subject to some of the constraints on arbitrage and risks that we have discussed already.
But why should such cross market anomalies arise in the first instance?
One possible explanation is that there is segmentation between the equity and options market – in other words, some players are confined to only one market. For example, a long-only equity investment fund may not be permitted to trade in options. Such constrained players could actually exacerbate pricing anomalies between markets, if forced to trade within a single market segment regardless of price because of, say, client redemptions. Furthermore, they would be unable to correct pricing anomalies between markets through arbitrage.
For traders and risk arbitrageurs, this suggests that cross-market flexibility is important for maximising the opportunity set.
Free money
Examples of cross market anomalies abound. A notable case involves the issuance on 31 October, 2008 by the British bank, Barclays plc, of £4.3 billion of new mandatory convertible notes (MCNs) in an effort to raise additional tier-one capital during the economic slowdown that year. The MCNs were issued at more than a 22% price discount to the ordinary shares and yet offered a higher yield and a fixed conversion price (subject to adjustment clauses for any future equity issuance below the MCN conversion price). The discount offered by the MCNs created a clear arbitrage opportunity against the equities of the firm.
There was also good liquidity in Barclays’ ordinary shares – the third panel of Figure 1.1 below shows that there was elevated trading volume in the ordinary shares on the day of issuance of the MCNs, with approximately £200 million traded. On 31 October, the firm’s ordinary shares initially rose sharply in price, from 205.25p to 217p as the market absorbed the early-morning announcement of the issuance. Only then did they start to fall as expected, ending the day down at 172.59p.
Figure 1.1 - Barclays PLC share price, relative share price and trading volume around 31 October 2008 MCN issue
Source: Thomson Reuters
From publicly issued documents and regulatory news-service releases it is clear that at least one long-short manager that held short positions in Barclays equity bought MCNs in scale (thus covering at least part of their short position at a favourable price), and at least one long-only manager sold existing long positions in ordinary shares and bought a large number of MCNs. Both of these were rational trading strategies. Nevertheless, many millions of arbitrage profits remained available for those able to exploit the opportunity – there was good liquidity in the ordinary shares at prices well above the offered MCN price and there were hours to make the required trades. Why would millions of arbitrage profits be left to collect for hours on end?
If $100 bills were dropped on the sidewalks of Chicago at eight o’clock in the morning, would you expect them to still be there by lunch time?
And yet this appeared to happen on the London Stock Exchange on 31 October 2008.
One possible explanation is that markets were segmented. That is, investors could not buy the MCNs by mandate, or at least were not sure if they could buy the MCNs (and could not get clarification from their compliance and legal teams quickly enough to take advantage of the open offer!).
Another potential explanation is short–selling constraints. About a month earlier, the UK Financial Service Authority had prohibited the active creation or increase of net-short positions in publicly quoted financial companies, although existing short positions were unaffected. This prohibited those traders without legacy short positions in Barclays ordinary shares from under-taking capital-structure arbitrage based around the issue of the MCNs. It is also possible that holders of existing short positions in Barclays’ ordinary shares feared that they might have been unable to hold on to these positions until conversion of the MCNs. Alternatively, they might have been unsure about how many MCNs they could obtain in the open offer (although this seems unlikely – there was a huge supply of MCNs on offer!). Finally, it could simply have been investor error: inertia or a misunderstanding of the nature of the MCNs.
Note from Figure 1.2 below that stock lending did not increase around the time of the MCN issue, consistent with the FSA ban on short-selling (although the increase in stock borrowing in early November is interesting!).
Figure 1.2 - Stock lending activity around 31 October 2008 MCN issue
Source: Data Explorers and Thomson Reuters
Two important lessons emerge for traders.
First, they should seek as much trading flexibility as possible (i.e. a broadly-worded mandate).
Secondly, they should approach any new capital instrument as an opportunity for capital-structure arbitrage; immediately reading the prospectus or similar documentation so as to understand the relationships within the new capital structure. Together, these should allow the trader to exploit mis-pricing opportunities that arise as a result of market segmentation.
Short-sale constraints
Earlier, we considered the notion that short-sale constraints matter in markets, as they can prevent traders from exploiting mis-pricings and can lead to security prices remaining over-valued for some time. A short-sale generally requires the borrowing of securities to facilitate the settlement of the transaction [3] . However, it is not always possible to locate securities for borrowing. Also, the short-seller must generally pay a fee to borrow securities, and this can reduce the attractiveness of the short-sale. These problems, plus legal barriers, are known as direct short-sale constraints.
There are also many indirect constraints on short-selling, including the potential for unlimited losses and the risk of being caught in a crowded exit. In extensive interviews that I conducted with short-sellers and prospective short-sellers [4] , interviewees identified no less than 34 barriers and difficulties with short-selling. There could even be more than this! These constraints tend to be risk-related, social or institutional in nature.
Perhaps one of the most interesting barriers mentioned is the perception that short-selling is a ‘trading’ activity rather than an ‘investing’ activity, so that it becomes unacceptable in the eyes of some stakeholders, such as trustees, consultants and ultimate clients. On the whole, our understanding of the risks associated with short-selling is limited. (I examine several of these in greater detail in the next few chapters.)
To what extent do short-sale constraints