3 Dec, 21 tweets, 5 min read
Trading is a game of repeatedly asking:

1. What exposures do I want?
2. What exposures do I have?
3. How do I get closer to 1 from 2, given that:
a) it costs to switch positions
b) my estimates are noisy
c) co-movements of assets are somewhat predictable

1/n
Concepts like "open trades" and "unrealized p&l" tend to be unhelpful in this paradigm.

If you don't like the exposures you have, then move them closer to the ones you want.

It makes no difference if you're underwater or in profit in your "position accounting"

2/n
There is no difference between a position that you have kept on the book for a while and one you just opened. It's exactly the same exposure either way.

Let's run through these 3 questions using a simple toy trading approach...

3/n
In our toy trading strategy, we have a universe of 3 assets, and each week we want to:
- go long the asset we think will have the highest returns the following week
- go short an equal \$ amount of the asset we think will have the lowest returns the following week

4/n
1. What Exposures Do I Want?

To answer this I need some *prediction* of the relative returns of the assets over the following week.

This might be from discretionary views. Or it may be a complex multi-factor model.

Either way, it needs to embrace uncertainty.

5/n
We don't know which asset will outperform next week.

So our prediction, visualized, might look something like this.

For each asset, we plot next week's return on the x-axis and (our prediction of) the probability of that asset experiencing that return on the y-axis.

6/n
The peaks show the expected return of each asset. That's the average return we think we'd get if we could observe next week's returns a huge number of times

(Maybe by rewinding the universe and replaying it again with a different butterfly wing flap)

7/n
But the overlap between the distributions shows us that lots of different outcomes could reasonably happen next week.

There's a reasonable chance that blue outperforms red next week, even though it's not the *expected* result.

8/n
If there were no overlap in the distributions then we'd be predicting that red was *certain* to outperform blue next week.

If you've traded at all, you know that's dumb.

9/n
Red outperforming blue is the *expected* outcome.

It won't necessarily happen this time but, if our predictions are good, then we'll make money over the long run taking bets like this.

So under our self-imposed trading rules, we want to:
- go long red
- go short blue

10/n
2. What Exposures Do I Have?

In this example, I'm just getting started trading.

I just have some cash.

11/n
3. How do I get closer to 1 from 2, given that:
a) it costs to switch positions
b) my estimates are noisy
c) co-movements of assets are somewhat predictable?

We just have cash. So it costs approximately the same amount to get into any asset exposure.

12/n
So I go straight into the exposures I want.

I buy x dollars of red and sell x dollars of blue.

Now market stuff happens...

My exposures change size in my portfolio and my predictions change.

So next week, the exposures I want and exposures I have are both different.

13/n
We ask the three questions again...

1. What exposures do I want?
2. What exposures do I have?
3. How do I get closer to 1 from 2, given that:
a) it costs to switch positions
b) my estimates are noisy
c) co-movements of assets are somewhat predictable?

14/n
1. What exposures do I want?

Let's say my predictions now look like this...

Blue still has the lowest expected returns.

Green now has the highest expected returns (by a small margin.)

So under ideal conditions, I'd want to be long green and short blue.

15/n
2. What exposures do I have?

I am long red and short blue.

I started off equal \$ in each one, but it will have got somewhat lopsided as prices move.

For simplicity, let's assume not too much.

16/n
3. How do I get closer to 1 from 2, given costs to switch positions, and uncertainty in estimates etc?

I can stay in my current exposures (long red, short blue) for free.

To switch exposures to the theoretical ideal ones requires me to sell red and buy green.

17/n
If I switch exposures I am paying *certain* trading costs, in exchange for an *uncertain* increase in expected return.

How I feel about that depends on:
- the cost to trade
- the expected increase in expect returns
- other effects such as asset covariance predictions

18/n
Likely, though, the small uncertain increase in expected returns is not a reasonable trade-off for paying the *certain* costs to switch position.

So I would keep current positions on.

19/n
If my predictions looked more like this, however, then I would pay up to move towards my theoretical preferred exposures.

The significant increase in expected return probably justifies paying up to switch from long red to long green.

20/n
So that's the pattern. Repeatedly ask the 3 questions.

1. What exposures do I want
2. What exposures do I have
3. How do I get closer to 1 from 2, given my constraints and models?

Doesn't matter how you got where you are, or how you make predictions. Ask the questions.

21/21

• • •

Missing some Tweet in this thread? You can try to force a refresh

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

# More from @therobotjames

2 Dec
In this brilliant article, Kris talks about the importance of trying to "shorten the feedback loop"?

How does a fundamental manager "shorten the feedback loop"?

I don't do anything that looks like "fundamental investing" but I couldn't stop thinking about the problem...

1/10
I think of the returns from "fundamental investing" coming from two sources...

1. Risk Premium - The tendency of risky assets to be relatively cheap vs their expected cashflows. This leads them to "carry" more than they would if their real cashflows were riskless

2/10
2. Mispricing - For behavioural/structural reasons, some assets are under/over-priced vs a reasonable estimate of their ex-ante "fair value".

On average, we expect them to converge towards fair value over some long time horizon.

This is "alpha" in fundamental investing

3/10
2 Dec
Traders often implicitly assume recent conditions will persist - without checking whether that's likely to be true.

You need to understand the assumptions you are making, and whether they are reasonable.

Here are some examples and a simple analytical approach 👇👇👇

1/n
"I want to find an asset with characteristics X

"So I'm going to look for stuff which showed these characteristics in the recent past - and I'll hope it carries on having them for a while."

Sometimes this is reasonable. Sometimes it is wishful thinking.

How do we tell?

2/n
Here's an example of something we know is persistent - volatility.

"I want to find a low volatility stock - so I'm going to look for stocks which have been low vol this year"

Is this reasonable, or wishful thinking? Let's see...

3/n
27 Nov
You think you've identified a new, useful predictive factor for trading...

But is it really new? Or just another way of looking at something you already know about?

How might you tell? Here are some simple ways...

A research thread 👇👇👇
First, put aside any expectation that you can isolate and quantify effects with great precision.

The market is a highly efficient beast - why means that any non-random effects we observe tend to be extremely noisy.

But just cos something is hard, doesn't mean we shouldn't try.
In fact, it's essential that we try to understand and isolate effects as best we can.

The best tools for the job (at least to start) are:
- economic intuition
- very simple data analysis (the kind of thing you could do in an excel pivot table)
12 Nov
Shall we do some analysis on a *really dumb* factor which might predict relative returns in stocks?

"Are cheap stocks expensive?"

A research thread 👇👇👇
Options on stocks with a low share price tend to be overpriced.

Equity options (at 100 shares a pop) are quite big for a small retail trader. So we might say there is excess retail demand for options on cheap stocks - which would result in them being overpriced.
But are low priced stocks also expensive?

The AMZN share price is \$3k+. There are Robinhooders who can't afford a single stock.

Do we see the same effect in Stocks as we do in the options?