New York City 5th Ave Vertical_edited_edited.jpg

Subscribers please log in to view the content

More buyers or sellers: It's a new dawn and I'm feline good

How to recognize a dead cat

The following is a true story. Names have not been changed, because, well you'll see...

At more buyers or sellers HQ, we recently acquired a stuffed dog, when our 2 year old grabbed it off the shelf at Ikea. It is a husky, an unfamiliar breed to him. A husky has pointy ears, a characteristic it shares with a different quadruped. As such, due to the wonders of a 2 year old mind, we now have a stuffed dog called Cat.

Cat and Bear were fortunate to survive the hit and run incident pictured here. If that is the size of a bear, you can see that scale doesn't factor much into the stuffed animal identification tree. And yes our kitchen floor really is that colour. The house was built in the 80s; what can I tell you?

Presumably demand for stuffed cats is limited, because if you want a fluffy cat-shaped object that will stay on your bed all day, then you can get a cat. Nevertheless, we mused over the possibility of sourcing a stuffed cat, and then calling it Doug, you know, because we're weird, and thoroughly committed to maximising comedy. It must be confusing to grow up in our house with all the nonsense that goes on.

Fast forward a couple of months, to the scene at a local playground. Picture if you will another 2 year old climbing up a six foot tall rock structure, spreading his arms and then leaping off like Icarus, to be caught by his father. Yes, that happened. Awed by this bravery, we wandered over to the swings where a Grandfather was pushing yet another 2 year old. On the floor was that rarest of finds: a stuffed cat; in this case a Jaguar. The Grandfather picked it up, handed it to his Grandson and said, 'Here you go, here's Doug.'


On the topic of non-living cats, last week we asserted without evidence that FB and AAPL were in the midst of a dead cat bounce. What did we mean by this, and other than the unreliable pointy ears, how would you recognize one when you see one? More generally, what price patterns do we find to be meaningful and useful?

Which side of a cat has more hairs?

This is an old joke that you may have heard.

An old joke that you almost certainly haven't heard is:

'Here have a banana', 'No thanks, I'll smoke it later'.

Somehow this seems funnier now than it did hearing it for the umpteenth time 30 years ago.

Anyway, for those that don't know, a cat has more hairs on the outside.

One of the most basic problems in topology is to prove that if you draw some kind of loop on a piece of paper, it has both an inside and an outside. This seems obvious, but if you twist the paper into a Möbius strip (pictured) it is no longer true, so maybe it's not that obvious. It depends on some global structural property of the paper.

The usual way to address such problems is to transform the complicated space into a simpler object. In the paper case this object is the group of integers, where we basically count how many times the loop wraps around a point on the inside.

We can think of the price series for a stock as being an example of such a transformation. We take an extremely complicated object that is a company, and convert it into series of positive real numbers. The price doesn't give us any insight as to why a company is doing well or poorly, but price patterns may reveal structural properties that can guide us.

Perhaps the most basic price pattern is a trend line, so let's start there.

On the catwalk

Cats have a remarkable ability to walk in a straight line along a narrow fence (and then leap 6 feet onto a roof to get in through an open bedroom window). Traders will similarly use straight lines to describe the path a stock is expected to follow. What structural factors might cause trend lines to exist?

One example we have discussed previously are the converging trend lines that form a symmetric triangle. We think these can arise as a result of fixed risk parameters. Specifically, it is reasonable to assume that an option market maker will manage their risk by flattening their delta whenever it reaches a certain fixed threshold. As a large option approaches its expiry, the distance that the stock has to move to reach this threshold decreases. Therefore, the market maker will act to reverse the path of the stock in a progressively tighter range.

Okay so that makes sense for some short term trend lines, but what about those long term trend lines that people draw?

As an abstract example, suppose we have a mysterious new cryptocurrency called the MöBuyOS. This single sided coin is backed by a fund of unknown value, but otherwise has no purpose other than to be a vehicle for lively discussion. If the number of comments received on a given day is even, the fund is incremented, if odd it is decremented. (Yes odd numbers are bad, we all know this.) Since this idea is stupid, quite often no comments are received, which gives the fund a small positive drift.

No-one has any idea what the value should be, so the price of MöBuyOS will wander about quite randomly, even accounting for the drift.

Suppose however that the entity that created the coin is allowed to buy a fixed amount each quarter if they wish, using cash from the fund. So long as the market price is below what they know the value to be, they will do so. This regular buying of a fixed amount will create a somewhat linear up-trend in the price.

This was a nonsense example, but it sounds a lot like a company doing regular buybacks.

Source: Yahoo

Since the first quarter of 2019, AAPL has conducted roughly $20bn of buybacks each quarter. (We saw this on the internet so it must be true.) It seems plausible that this could be responsible for two trend lines of roughly equal slope during this period. The boost that tech companies received from the pandemic explains the outperformance relative to these trends in 2020.

If the price fails to hold the current trend line around 135, we might reasonably infer that either buybacks have declined substantially, or that the fundamentals have deteriorated. In either case a larger downtrend is possible.

Source: Yahoo

There are other reasons why long term trend lines might appear. For example, in FX, a difference in the inflation rate between two countries is likely to create a trend between their respective currencies.

Any others you can think of? We'd love to hear about them (in pairs).

On a much shorter time scale, a TWAP algorithm is another example of regular buying that will create a trend line. This is not a very easy thing to take advantage of, but it is sometimes useful to know that a change in price was due to an order as opposed to a change in fundamentals.

Schrödinger's cat