A+? AA1? BB-? It seems as though company ratings have been here forever. Market leaders S&P, Moody’s & Fitch have pretty much written the ultimate rating book. But the times are now changing, and fast, and that manual is becoming increasingly incomplete, perhaps even irrelevant.
Present day company ratings are focused on determining a company’s ability to pay back debt. Short term, long term, but that’s pretty much it.
This is important, say, if you’re considering giving the company a loan. If you wish to partner with the company, long it, short it, interview for a job or buy its products, this specific rating says too little.
In addition, today’s rating practice is based on the analysis of financial & market reports. Since private companies are not obligated to publish their financial information, these reports usually apply only to public companies.
There are a lot of public companies — 100,000 globally, more or less and decreasing— but there are many millions of private companies, and these cannot be effectively rated this way.
The process of analyzing financial and market reports is labor intensive and manual. New technologies are being developed for the purpose of scalable analysis, but for the most part, the rating industry is rating yesterday’s news using very expensive people.
The internet has blessed us with an unprecedented array of potential rating ingredients. Google, Linkedin, Alexa, Facebook, Twitter, IP & Trademarkdirectories, to name a few, tell us so much about companies, private and public alike, it makes no sense to not use them.
These signals are much harder to manipulate and just as importantly, they appear as they happen, in real time.
So how should we use them?
Here’s a simple list of signal categories and how they can be used to rate any company, private or public:
- Competition: Who are the company’s real competitors? In what aspects is the company better or worse? Bigger or smaller? Present or absent? Proper analysis of all news, publications, and reviews mentioning the company can draw a pretty accurate set of answers to these question.
- Team: How deep is the team? How long have they been together? What is the ratio of executives to people who actually work? Of production employees to accountants? Is the team growing? What is the average tenure? How do they fare against competitors? All and more can be found with proper use of Linkedin, for example.
- Market Acceptance: Is customer traffic increasing? Organically? What kind of customers? Do they appreciate the products? The company? Reviews and traffic management monitors can provide a good start. Cunning use of Google can do most of the rest.
- Sentiment: Is the company’s momentum on a positive or negative trajectory? Is the company about the burst through glass ceilings or is shit about to hit the fan? Most of the time Google can provide quite a few hints and so do Linkedin employee profiles and Facebook / Twitter rants.
- Depth: Does the company possess IP and Trademarks that are both unique and relevant to its future? IP directories have that information. It just needs to be processed and, more importantly, compared with the competitors and the rest of the relevant market.
- Financial and Shareholder Environment: Companies tend to advertise good things, and solid investors are usually a good thing that can be easily found. Companies also tend to publish good financial results and customer wins, and while this is not a stand-alone indicator for anything, its inclusion in a wide array of signals can tilt the scale from B+ to A- and vice versa.
There are many additional signals and countless ways of processing. But, there are challenges as well, the main ones being:
Collecting all this info in real time, not to mention across markets, geographies and languages is f*&%ing difficult. It requires quite a bit of thinking, planning, engineering, testing and calibrating.
Current rating methods have been around for many decades. Using internet signals would require developing a new methodology and testing that methodology across markets, time frames and false positives/negatives. It would also require reconsidering rating fundamentals, such as stand-alone vs. relative and the effectiveness of the letter scale.
Depth & Data
There’s a reason why ratings have become what they are today. Years of data collection and calibration have actually contributed to a formula that works ok. Yes, it can be better, but it can also be worse. The change is expensive, the risk can be high if mismanaged.
This is, perhaps, the most significant challenge. Rating, simply put, is a profitable business.
Proper implementation of a new rating system will involve managing a delicate balance of old and new, solid experience vs. necessary innovation, risk avoiding and chance taking, proving technology vs. experimental. This, perhaps, will be the biggest challenge the winner of this market will need to master.
All in all, these are good reasons to take things slow, except that one’s challenge is another’s opportunity.
The rating universe is quietly undergoing a significant and disruptive evolution process. AI and Data companies are creating the foundation for new systems, and existing market leaders have two options: adapt or die.
Since a bad rating system presents a significant risk to everyone involved, the preferred scenario is of evolution and a synergetic process that will merge between the new and existing systems.
That said, the technology is developing rapidly — lack of evolution will likely result in a revolution.
Or, as Billy Bean cleverly noted: adapt or die.