Fred’s Views - Issue 8
A weekly summary of what I've found interesting at the intersection of economics, finance and technology.
Keeping with the theme of corporate sources of funding in Asia (see Issue 7), the new Star Market in Shanghai, focusing on technology companies, started trading last week (FT $). The authorities’ stated goal was to increase the maturity of capital markets, “to ensure that the right funding reaches the right companies at the right price.” There were restrictions on who could trade on this new board, with minimum funds and experience requirements for example, but in the end the pent-up demand for a piece of the action in China's tech industry proved overwhelming, and shares popped on the opening day. Of course now the question is whether the new board will succeed in becoming a source of long-term funding, or whether it will just be another venue for speculation.
Journalists and other experts investigating Russian intelligence activities were targeted by sophisticated phishing campaigns (FT $). The interesting wrinkle is that the targets are using ProtonMail, one of the world’s best-known secure email services. The targets of the campaign are obviously sophisticated users, and yet it seems as if a breach may have occurred, which serves as a good reminder that the human link is often the weakest in the security chain, or at least the easiest to exploit.
In related news, a German public broadcaster did an investigation into incidences of cyber espionage (BR), and discovered that nearly all German corporations in the DAX index, often leaders in their field, were compromised to a lesser or greater degree. The interesting angle here is that the attacks were not particularly sophisticated, but the sheer volume of them nearly guarantees that some valuable information will be uncovered. Another reminder of the stark asymmetry in costs for trying to secure a company vs trying to compromise it.
However, most people shouldn't worry too much about such attacks, and more about more mundane pests like adware (Wired)
Amazon will try everything in the retail industry's 150-year old playbook, but they haven't yet cracked the overall shopping experience. Read (Wired)
Profile of Morgan Beller, the driving force behind Facebook's Libra. Read (CNBC)
Sequoia Capital is turning its focus to South and Southeast Asia to find future unicorns, after seeing a decline in opportunities in China, a result of their longtime presence there as well as the huge amounts of funding available in that market. Read (FT $)
The London Stock Exchange is in talks to buy Refinitiv, which used to be Thomson Reuters’ market data business. If it succeeds, it could emerge as a credible rival to Bloomberg. It's established that markets data is one of the few types of data that people are willing to pay for, but I wonder whether this is not a merger rooted in the past, rather than the future... Read (Business Insider)
Payments looks like the financial services function most likely to be disrupted by tech upstarts. It's a technological problem to start with and the current solutions leave to be desired (slow, cumbersome). But incumbents such as SWIFT may well be best placed to successfully disrupt themselves. Case in point: in a recent trial, SWIFT managed to bring down the processing time for cross-border payments in Asia to just seconds, using its new gpi platform. Read (Swift)
WeWork is planning to IPO as early as September. Cashing out while they can? Read (FT $)
Ride-hailing cars account for nearly one third of Manhattan traffic, and traffic speeds have fallen almost 30% compared to 2010. So much for increasing efficiency on the road... Read (Twitter)
Very interesting look at the realities behind AI and data, and a related framework for analysis, applied to the US vs China. Read (MacroPolo)
A popular ML algorithm used to identify “toxic” speech on social media platforms, rates tweets by far-right extremists as less toxic than those of drag queens, who inevitably are usually the target of said toxicity. Sad confirmation that context entirely eludes such algorithms, and that by extension they are not the solution to policing behaviour online. Read (InternetLab)
This is not a surprise, obviously, but Apple also listens to Siri recordings as part of its quality control processes, and contractors end up hearing things they probably shouldn't have... (see Issue 6 for a similar story with Google Assistant) Read (The Guardian)
It's fairly trivial to draw up a detailed map of a random Instagram user’s whereabouts, using the location data they willingly uploaded to the system. See also the danger of the human factor higher up in this newsletter. Read (Wired)
That's it for this week's edition. As always, thanks for reading and please forward this to anyone who you think might be interested, it would be much appreciated.