Précédent Suivant

11. Establishing Robust Markets for Infrastructure-backed Securities

p. 75-79


Texte intégral

1In addition to capital raising through bank lending and primary bond and stock offerings, a robust global financial architecture must provide the opportunity for institutional investors to shift portfolio profiles easily, quickly, and cheaply in the face of change. Such changes include interest rate expectations, risk perceptions, comparative returns, and other aspects of the portfolio optimization search. And because most institutional investors are fiduciaries, they are obliged to act in their beneficiaries’ interests — that is, institutional investors are subject to both “duty of care” and “duty of loyalty” to their clients.

2The exercise of fund management obligations may be made more difficult in the case of infrastructure financings because of the nature of the financial instruments that emanate from them. This may be a key bottleneck for financial efficiency in the infrastructure “air supply” mentioned earlier. Consequently, the focus has to be on the efficiency of both the primary and secondary markets for debt instruments generated by project and infrastructure financings.

11.1 Primary Markets for Infrastructure-backed Securities

3A “primary” market for a security provides pricing and transfer of ownership from the issuer to the initial holders. A “secondary” market allows for subsequent purchases and sales that do not involve the issuer. Assuming that an infrastructure-backed security (IBS) will closely resemble traditional debt and equity securities, we can sketch out how these markets might operate.

4The primary market needs to be structured so as to encourage private production of information and competition among potential investors. Information production is necessary because IBS valuation will require specialized knowledge given the underlying projects’ scale and complexity, even under the most thorough and transparent disclosure regimes. Vigorous competition among investors, transparency, and market discipline will presumably address these concerns for purely private-sector issues. Even with public-sector issues, competition is necessary because the governments likely to be the securities’ primary guarantors need to convince their diverse political constituencies that the project is not a “giveaway” to wealthy and well-connected investors.

5The two standard frameworks for primary markets are auctions and underwriting syndicates.

6Auctions (such as the “modified Dutch auction” used in Google’s 2004 IPO, or the primary dealer-based auctions used for US Treasury securities) are generally viewed as transparent, fair, and competitive. Competition in auctions, however, depends on high participation (many bidders) as well as measures to discourage collusion. The record of electromagnetic spectrum auctions, which can be viewed as large direct transfers of infrastructure ownership, has been notably flawed in this regard.1

7Moreover, even though some observers expected Google’s IPO to validate the auction format and lead to widespread adoption of auction procedures, the number of equity IPO auctions has not seen strong growth. Research has suggested that the high efficiency generally associated with auctions actually allows too much free-riding on the costly production of private information. This argument can be used to justify the conventional IPO underwriting process, despite its high cost in terms of fees and persistent underpricing.

8In most countries today, underwriting syndicates remain the dominant primary market mechanism for issuing risky securities.2 Because production of private information is likely to be a key element of the infrastructure-backed securities origination process in the debt capital markets, similar considerations will likely apply to this sector as well.

11.2 Secondary Markets for Infrastructure-backed Securities

9Although a viable secondary market is not strictly essential for infrastructure finance, future opportunities for IBS resale can increase the securities’ value in the primary market, which will reduce borrowing rates and therefore the cost of capital.

10IBS are inherently long-term securities, as discussed earlier — longer than the investment horizons of all big asset pools except perhaps endowment funds and multi-generation family offices. IBS are also undiversified. Both of these dimensions suggest strong non-informational reasons for secondary-market trading.

11The most liquid secondary markets are for equities, and these are generally organized as electronic “limit order books”. Using this mechanism, traders enter buy and sell orders with limit prices. If a new incoming buy order’s limit price meets or exceeds the lowest price of the previously entered sell orders (the “ask book”), the orders are matched and a trade occurs. If the incoming buy order cannot be matched against any sell order in the ask book, it is added to the bid book, where it will be available for matching against newly arriving sell orders.

12In the more widely used “lit” form of this market, the bid and ask books are visible and widely disseminated; in a “dark” market, bids and asks are not displayed. Because orders are handled continuously on a first-come, first-served basis, faster traders have an advantage over slower ones, which has given rise to the practice of automated high-frequency trading. The electronic limit order markets feature high transparency and are readily accessible by retail and institutional investors. This form of market organization dominates in equities and standardized futures and options contracts.

13These markets usually function well when there is sufficient natural trading interest, which is often associated with dispersed ownership and inclusion of the securities in an index or exchange-traded funds (ETFs). They sometimes require assistance, however, from designated market makers (DMMs) and auctions. DMMs assume the responsibility of ensuring that there is always a posted bid and offer during regular continuous trading sessions. When natural trading interest is too low to warrant operation of a continuous market, once-a-day “double auctions” are often used.

14Existing debt securities, such as bonds and notes, generally trade in dealer markets. These markets rely on intermediaries that take the other side of customer trades. Banks have traditionally performed this function. After the global financial crisis and application “Volcker Rule” as part of the US Dodd-Frank legislation, however, bans on proprietary trading have limited banks’ dealing capabilities. This limitation creates an opening for other players (e.g., hedge funds) that are willing to commit capital and develop expertise in the securities and markets concerned.3

15The quality of dealer markets in debt securities varies widely. They are generally viewed as adequately serving institutional trading needs for corporate and sovereign bonds. They are less successful in providing trading opportunities for retail investors. In contrast to equity markets, proportional trading costs are lower for larger institutional trades than smaller retail trades.

16For IBS, although the amount of capital raised in the retail sector might be small, the possibility of some local retail participation could help sustain consensus in the political sphere. Even at their best, dealer markets tend to suffer from low transparency.

17In some respects, infrastructure-backed debt securities may most closely resemble the existing municipal bond market in the US. This market segment includes debt that is often issued to build public infrastructure or, as in the case of IRBs, private infrastructure that serves an identifiable public purpose.

18The current municipal bond market does not provide a desirable template or starting point for an infrastructure finance architecture. It is extremely fragmented. The Mergent Online database lists about 3.6 million issues of municipal securities, roughly 10 times the number of US corporate debt issues. Trading costs are high. Sirri (2014) finds that the median customer-to-customer differential (similar to the bid-ask spread) is 198 bps. The average differential for retail-sized trades (up to $5,000) is 246 bps.

19The US municipal bond market therefore provides a cautionary model for IBS debt. In most cases, this model would be hampered by the absence of the interest tax subsidy that exists for municipal debt in the US.

20Finally, as noted earlier, the vast bulk of debt issued by municipalities and IRBs has final maturities of less than 10 years, whereas the financial time-profile of optimal debt for infrastructure projects often significantly exceeds that maturity. This dynamic suggests an extraordinary opportunity for improving capital market access for infrastructure projects by creating new ways of generating liquidity for investors.

Notes de bas de page

1 See Binmore and Klemperer (2002) and Klemperer (2002a, 2002b).

2 See Sherman (2005) and Chiang, Qian, and Sherman (2010).

3 Whitehead (2011).

Précédent Suivant

Le texte seul est utilisable sous licence Creative Commons - Attribution 4.0 International - CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.