Capital Project Financing, or Public Investment Management Financing
There has been a push for local, or subnational, governments in the developing world to issue bonds to supplement financing for capital projects. In the US, it is pointed out, municipally issued bonds have been a stable market for over a century and are used to provide the upfront cash for most local infrastructure, and even for social impact bonds. Why not use this mechanism in the developing world?
There are several basic questions. First, one must not think of what you as the end user of the cash want. You are asking someone else, bankers or even a broader general public of buyers, to lend you cash. As a former municipal public finance investment banker in the USA (the market people usually refer to), and now consulting internationally, let me bring to the discussion some of these basic questions.
Question 1 –What information does a lender need? For municipal bonds, this would be measures of their financial performance (income statements) and current financial situation or position (balance sheets). For a social impact bond they are looking at progress (typically milestones against a certain baseline) on meeting certain social or public service needs. These could be the number of acres of trees planted (for carbon offsets, the subject of the original article), number of children that have received education in a year, reduction in levels of sickness, increase in number of small businesses established, etc. To fund proposed ventures, this would be measures of the amount of business available (size of the market, number of competitors), and the ability of the project to deliver (not only the ability and competence of the team, but can the infrastructure and environment support them, with such basic elements as power supply, etc.).
Question 2 – How much data is out there, and is it reliable? Big data has provided a revolution in analysis fundamentally because of the following. It has been able to gather many thousands (or millions) of data points from electronic transactions (purchases, etc.) or electronic interactions (e.g. people clicking on items on social media, providing information on their choices). As we all know from our statistics classes. if we only obtain a small sample of data/survey on an activity, we may obtain a very distorted view or picture of what is going on. A large, randomly chosen sample set (which many thousands or millions of transactions will hopefully provide) is likely to give us that. Data that is inaccurate, or distorted, is likely to be offset in such a large pool. There is another significant boon with Big Data. Each item examined may in and of itself give us only a small piece of information (e.g. the fact that a user clicks on one advertisement may not mean they really like that type of product, but it might show a small preference). But with so much data available now for so many items, an analyst can examine a large number of variables, and a large sample set for each of those variables.
Where is this data being produced? Credit card transactions, clicks on social media sites (Facebook, etc.) thousands and millions of times a day. Teraflops of data. But is the data that lenders need being produced in this way? Is the information that lenders for social impact bonds, or to governments in developing countries, being produced in mass fashion? Typically, no. The data that is needed are individual pieces of information that one or a few people will produce in a report or in a few transactions. These may or not be verified before they are entered into any system. In the case of government records, whether accounting records for their own transactions or records of property, etc. the completeness, accuracy and timeliness of records are typically dismal. I have spent the last decade assisting governments in Asia and Africa to improve upon the most rudimentary processes, and I can tell you that I would not put trust my money based on the information that is to come out of there. Social impact bonds are relying on data collected by specified individuals, and specific M&E systems.
Those systems in turn are reliant on a backdrop of a profession of auditors and related professions, and a culture of accountability with penalties for misreporting. Laws and statutes have to be in place, and professional codes of conduct have to be in place. In western countries, this social infrastructure has been developing over the last 150 years, and still has significant issues (the Enron scandal of 15 years ago brought down one of the Big 5 audit firms, as they were willing to agree to fabrication in exchange for earning large consulting fees). If data is unreliable, it must be examined even more closely, by analysts, not through a formula that is only looking at large scale patterns and has to ignore the important details that can highlight what may be wrong or right with the data.
By Gregg Pavitt, President, Pavitt Public Finance https://pfmexpert.net Twitter pfmexpert