Robot James 🤖🏖 Profile picture
Feb 10, 2021 22 tweets 4 min read Read on X
Broadly, there are 3 types of systematic trading strategy that can "work".

In order of increasing turnover:
1. Risk premia harvesting
2. Economically-sensible, statistically-quantifiable slow-converging inefficiencies
3. Trading fast-converging supply/demand imbalances

👇👇👇
1. Risk Premia Harvesting - is typically the domain of wealth management, but it's important to any trader who likes money.

A "risk premium" is the excess return you might expect over and above risk-free cashflows for taking on certain unattractive risks
"Equity Risk Premium", for example, is how you say "Stonks, they go up" if you work for Blackrock

(Though, they're really referring to the extent to which they "go up" more than an equivalent less risky thing)
The basics of risk premia harvesting are:
- intentionally expose your portfolio to diverse sources of risk that tend to be rewarded
- manage risk sensibly so no risk dominates at any time
- be patient and chill the f out (hardest bit for most)
A very simple example is the 60/40 stock/bond portfolio.

More balanced implementations include Bridgewater's "All Weather" portfolio and "Risk Parity" strategies, generally, which attempt to equalize risk across assets or risk premia.
"Doing useful things" like providing liquidity in a highly stressed market - or making a 2 sided market at all times are also, arguably "risk premia harvesting".

(You're taking on the risk of getting run over by those who can choose when to trade)
Risk Premia Harvesting is something nearly every trader should do.

All active traders need to be aware of it too

Want to short stocks? You're facing a big hurdle. You need to be right over and above the expected drift in the asset

It's better to play easy games than hard ones
Now the 2nd category: Economically-sensible, statistically-quantifiable slow-converging inefficiencies.

These are noisy tendencies for assets to trade too cheap/expensive at certain times due to behavioural or structural effects.
Examples include momentum effects, seasonal regularities, effects due to indexing inclusion/exclusion.

We can probably lump "style factors" (momentum, value, carry, quality, low vol) and most medium freq "stat arb" approaches in this bucket too.
This stuff tends to be noisy and "slow converging".

It's just a noisy tendency for "cheap stuff to outperform expensive stuff" (say)

So you have to analyze it "in aggregate" over large data sets.

It also means that your P&L is very slow to converge to "expected returns"
So to trade this stuff effectively we need:
- to understand why the inefficiency would persist
- faster converging metrics around what we're exploiting so we're not "the last to know" when the inefficiency disappears
- patience and discipline to "keep swinging the bat"
Useful sources for this kind of stuff include:
- Expected Returns, Antti Ilmanen
- Efficiently Inefficient - Pedersen
- Positional Option Trading @SinclairEuan (which literally gives you stuff to trade)
- Active Portfolio Management - Grinhold, Kahn
Generally, this stuff is less reliable than 1) risk premia harvesting and 3) fast-converging flow effects.

"Home gamer" traders usually spend too much time here, and too little time on risk premia harvesting.

Active managers of size play here tho they'd rather be playing 3)
3rd category: Trading fast-converging supply-demand imbalances

This stuff is the bread-and-butter of proprietary trading firms.

Short term supply/demand imbalances create dislocations in prices which fast traders can "disperse" by trading against and offsetting risk elsewhere.
These trades are conceptually simple and economically sound.

For example, I might buy sell futures on Shanghai INE and buy a similar contract cheaper on Singapore SGX for a profit (after costs).

That's a simple "arb" though it carries risks cos we can't trade instantaneously
More normally though, we're doing riskier trades that we expect to work out on average.

We're serially looking to buy cheap and sell high based usually on simple relative-value models.

The assumption is made that deviations from (relative) fair value will converge...
So models are less about "predicting the future" (as per 2) and more about "extrapolating the present" (Q vs P).

"But you're *predicting* convergence to your model of fair value you're using to quantify cheap/expensive?"

Yeah, exactly.
Advantages of these trades are:
- they're easy to understand and economically simple
- they converge fast to expected pnl. You know quickly when your model is out or you don't have edge anymore. They fit nicely with @KrisAbdelmessih's "measurement and normalization" paradigm
Disadvantages are that they are capital constrained (you can only eat what you are fed) and require significant investment in infrastructure and staff.

Whilst this area isn't practical for "home gamers", the lessons here are crucial for a good understanding of the market.
Takeaways for "home gamers"

- Do more of 1) and less of 2)
- Dig into 3) to understand and appreciate "the terrifying efficiency of the markets"
Here's a "work in progress" model of how 3 and 2 interact - and how you can do 2) by "sitting roughly in the right place with your mouth open"

this thread is now on my personal website here: robotjames.com/posts/three-ty…

• • •

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

Keep Current with Robot James 🤖🏖

Robot James 🤖🏖 Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

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

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @therobotjames

Jul 30
a chat today reminded me that the crucial first step in any successful trader’s journey is to…

stop doing really dumb shit.
if you have no edge (and i think we can both assume you won’t at the start) then there’s nowhere for returns to come from.

you can’t make money like that

but there are plenty of ways you can lose money.
1) if you have no edge then every trading approach apart from doing nothing can be expected to lose money.

trading costs money (from fees, spread, and the price impact of your own trades.)
Read 10 tweets
Jul 8
one of the most important things i tell people over and over again, like a stuck record, is that their trading should look like a useful thing that sucks.
you know that there are extremely sophisticated trading firms out there with ultra-low latency infrastructure and sophisticated modeling techniques.

and you might reasonably ask how you, as an individual, could possibly compete with that.
and the answer is that you can’t.

but you don’t have to.

you shouldn't even try.

so, why then, can many small speculators do ok and make money?
Read 9 tweets
May 21
nearly everything that is a good repeatable trading idea looks like:

"under <some circumstances> this thing is likely to be too cheap/rich because <some people> are being forced or greedy or stupid... so the thing is more likely to go up/down in the future" Image
your job as trader, operating in an efficient, competitive market, is to tell yourself that your idea about that is probably bullshit.

and quickly prove to yourself that it is indeed bullshit.

destroy those hopes and dreams quickly... and move onto something more productive. Image
you can show that something is a BAD idea way quicker than you can show yourself that it's a good idea.

and showing yourself quickly that something is a bad idea is a GOOD thing...
Read 16 tweets
Sep 30, 2024
all active etfs are trash.

under the premise that all active etfs are trash, i looked at what it would look like if you could shorta bunch of them against an equivalent SPY long.

the legs are sized to equal volatility based on 120 day rolling realized vol. Image
highlighly scientifically, i looked at etfdb and picked 15 active / tactical ETFs based on their name and category. Image
here's the performance of the long SPY / short ETF pairs individually.

some did less bad than others, but all the ETFs underperformed SPY, risk-adjusted.

FIG, HFND, MOOD look especially bad. Image
Read 6 tweets
May 17, 2024
andy's top didn't last all year, but it lasted 32 days.

is that a lot or a little?

it's a lot

if you called a top on every new 252-day high, most of the time, the call would fail the next day

the expected length a top would have held is 9 days

andy's top is 95% percentile Image
that the median case is to fail straight away should be self-evident.

if the market was 50/50 up or down on a given day, half of the time the top call would fail the next day.

but, as you know, the market prefers to go up, so the most common outcome is it failing the next day.
the mean of 9 days is pushed up by a few very long tops - such as the 1375 day one that started in october 2007.

here's what the histogram would look like if i didn't truncate the x-axis Image
Read 7 tweets
Apr 30, 2024
i think people new to markets massively underestimate how noisy everything is.

your job as trader is to try to work out when stuff is likely to go up or down, right?

then you can bet.

any trade might not make money but do enough good trades and you're likely to over time.
the problem you have, is that things go up or down for a million different reasons.

and the massive majority of those reasons are unknowable before they happen.

why?

cos tons of people are betting on this stuff, so all the obvious stuff gets priced in beforehand.
if we know something is gonna be trading $100 tomorrow, where's it trading today?

well, $100, give or take.

it trades for the price where you can't make any money trading on obvious shit everyone knows, right?
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(