Archive

Monthly Archives: November 2015

I’ve added to my DX Group position (London Stamps) after it fell more than 70% on the day of a profit warning. The news seems that the business is not doing as well as planned but it’s in no way catastrophic. It was already pretty modestly priced.

It’s now priced for a high chance of total failure, which seems an excessive reaction. Maybe the news being published during the trading day rather than before the open as usual triggered some weird conditions. Looking at the trades log on the London Stock Exchange website, it seems only £30K worth of trading in a post-announcement auction (uncrossing after 6 minutes post-announcement break) explains the bulk of the gap (80 to 55), and then the post-gap momentum may get its own life.

Usually things get worse before they get better after a profit warning (the earnings announcement drift anomaly) but in this case I’ve topped up on the day in case there’s a rebound early next week as bottom fishing value investors go back to their desks. I’m happy to keep the stock until it succeeds or fails more convincingly in any case.

Advertisements

Funny news about a granny who shred her cash stash to annoy her heirs. The local central bank will replace the banknotes!

Mattress safety

This bring interesting potential applications. How to make cash under the mattress safer? Just cut it in two and hide it under TWO mattresses in different locations. If you need to spend it, rejoin the two parts and redeem it for spendable currency at the central bank. (This doesn’t though cover the somehow unlikely scenario where the central bank ceases to operate but the banknotes still have some value.)

Paper multi-sig

This could also be used by a group of people who want to share a stash and only spend it when they all agree: cut each banknote in N-pieces where each party takes one. To spend, the parties must collaborate to reunite the parts and redeem them at the central bank. This is pretty similar to multi-sig addresses in Bitcoin — indeed it could be used to explain multi-sig to people who don’t get the explanation in cryptographic terms.

Shredded will

Back to the granny, a central bank representative is quoted as saying that “If we didn’t pay out the money then we would be punishing the wrong people.” I think it is the wrong attitude. Destroying a banknote is equivalent to making a donation to society, as the value of everyone’s remaining currency increases (there are fewer of them going around). From an accounting viewpoint it decreases the central bank, thus the state, liabilities as there are fewer banknotes to redeem (paper money in circulation show up as a liability on the issuer’s balance sheet). So, the act of shredding the notes was obviously a wilful act of spending on the granny’s side. She could have equivalently bought an expensive object and destroyed it so as to destroy its value.

Technicalities

The ECB says that they do not redeem banknotes damaged intentionally. The Bank of England seems more tolerant. Arguable if people started to use the procedure for mattress safety and paper multi-sig en masse, central banks’ damaged notes redeeming service could be overwhelmed, though economically or morally I don’t see a problem with operating the service, perhaps with a fee to cover direct costs.

Today’s trade in Obliquity London is swapping SAB Miller for Rolls Royce. The brewer got taken over and I don’t like it much for both fundamental reasons — big takeover tends to be underwhelming, and AB InDev’s management style is a bit too Ryanair-style for my taste — and practical — the new company won’t be London-listed.

Rolls Royce, the engine manufacturer, had been on the watch list for a while. It’s a pretty steady stock which is currently seen as more distressed than it really is in my view, and today’s profit warning may provide an interesting entry point.

Here’s an app idea: make an app, let’s call it Ubercut, that allows Uber users to shortcut Uber and avoid paying the fat commission.

How would it work?

The app would run in the background on a phone and detect if the regular Uber app is running in the foreground. When it is and both a driver and a punter are using it at the same time, and are within pickup distance, it sends a notification of the type “you guys may want to make a deal” to both parties, e.g. doing the ride but for 90% of the fare suggested by Uber (splitting Uber’s 20% comm) and allows them to communicate, thus bypassing Uber before initiating a transaction. A comprehensive hijack would “read over” both user’s Uber screens to send notifications only when there is an exact match. This could be quite tricky or impossible to implement, but a simple passive version (“running the app at the same time and near in space”) can be done relatively simply, at least in Android.

The contracting parties would lose the benefit of the protection afforded by Uber for on-platform rides, but most such platforms mainly work hard on their disclaimers when it comes to cover unwanted trouble. Third party insurers could also offer cover the weary, for a fraction of an Uber commission. The Uber reputation system would still be operational as this is pure freeloading — no ride are initiated in the app directly (in the simplest version of the concept, one could also couple it with a cheaper or peer-to-peer matching platform).

What’s the benefit?

Disruption! Uber and other trading platforms that are dominant collect economy rent due above the cost of operating the service because of the network effect — it’s a winner take all game. This is a classic market failure. The existence of shortcuts would probably contribute to cap such rents.

Would it be legal?

If Uber lawyers haven’t anticipated that, we can be sure they’d add a ban to their terms if such an app becomes popular enough for them to notice. Would people care? Maybe, maybe not. Could they could sue their clients and drivers succesfully? I know not.

Would it be moral?

Unambiguously yes I’d say, as it is merely about applying Uber’s philosophy (disrupt incumbents, ignore the law, etc) to Uber itself.