An introduction to quantitative investing

Our next speaker is Hugh Fram from Fram Funds.

Hugh established Fram Funds in 2016, which is an independent boutique investment manager who navigates financial markets using quantitative strategies and systematic risk management processes.

Fram Funds currently operates two funds, the Longshore Australian Equity Fund and the Fram Futures Fund.

Hugh is going to give us an introduction to quantitative investing.

Over to you, Hugh.

Great.

Thanks, Angeline.

I'll crack straight into it because I am aware of the time.

I'm going to be presenting something slightly different today.

Just it's going to be on quantitative investing, which is what my business specializes in.

We obviously have information about our strategies on there.

At the end of the presentation, there'll be the information there if you'd like to reach out so that we can talk to you in a little bit more detail about our funds and what we actually do.

Usual disclaimer that everybody knows, Frame Funds as a business, we are founder-led.

We focus on returns firstly, and we co-invest in all of our products.

We run two divisions, managed accounts, where we have two strategies, large cap Australian shares and also mid cap Australian shares.

And we have two managed funds.

One is a multi-track quantitative hedge fund, which trades internationally.

And then there is also quantitative Australian equities.

As a business, our objective is to compound investor capital at attractive rates over the long term, focusing on intergenerational funds management.

And we do this using quantitative and systematic investment strategies.

We have two funds and two managed account portfolios, large-cap Australian shares, mid-cap Australian shares, Fray Futures Fund, which trades internationally, and Longshore Australian Equity Fund, which is the quantitative Australian shares.

So firstly, when we get into it, what is quantitative investing?

It sounds very complicated on the surface and some parts of it is very complicated, but we'll try and make it as simple as possible and we'll give you a couple of examples.

We essentially, we use mathematical and statistical methods to design and execute investment strategies rather than relying on the opinion of a past experience or the opinion of a fund manager.

We attempt essentially to quantify all inputs

as part of coming up with an idea to invest.

We use historical data as a core evidence to build the strategy and by doing this we can use extensive mathematical and statistical testing

to figure out exactly what our assumptions are about the strategy.

This puts us in a position of flying less blind, essentially not having to sit here and go, I don't really know what to expect from this portfolio.

We can get some very clear assumptions about how this portfolio will trade in good times, bad times, and the type of volatility that it will undertake.

Ultimately, we are hunting for statistically significant repeating occurrences in financial markets.

We want to know when we do and build a strategy, we want to know what to buy, when to buy it, how much we're going to buy, the probability that this position is going to be successful, how often we can exploit this occurrence and the risk that we need to take on to take this trade and to make money out of it.

There are a few strategy types.

There is a normal distribution or a symmetric distribution curve.

We really try to target a combination of symmetric as well as right skewed distribution curves where we are with the right skew.

We're essentially trying to leave the tail open, where what we're trying to target is invested businesses that have the ability to multiply and make 10

20 times our money whilst managing the downside risk.

Some of the strategy examples that we implement within our portfolios, relative momentum, which is versus another, active momentum, which is versus a benchmark, mean reversion, trend following, and market structure.

Market structure is focusing on ETF rebalancing and where money flow is actually

going to and from.

There are many parts of our strategy development focus.

There are really four key ones which we'll briefly touch on.

Idea generation, which is the primary phase where we come up with a hypothesis of what we are wanting to build as a strategy and the type of distribution it will generate over a period of time.

Edge identification is refining that investment edge.

We figure out what type of return profile, what type of

how often it trades and then ways that we can improve the strategy through parameter optimisation.

We then move on to the next stage, which is robustness testing.

Some things people will have heard of, in and out of sample, Monte Carlo simulations, which is something that recently on LinkedIn we've been talking about a fair bit, breadth of edge, how dense and how wide is the edge or the probability that you're trying to identify and capitalise on.

You go through signal randomisation, cross-asset testing, which essentially would be testing between commodity markets as well as fixed income markets, as well as equity markets, as well as currency markets.

Transaction cost analysis and noise and stress testing are the things that we've also spoken about recently through LinkedIn and written articles on.

When we are about to implement, we need to do some pre-implementation testing as well if infrastructure

forward testing and also we like to do back testing and forward test analysis to see how accurate it is.

After we've done this, there are other parts that we look to do as well, is post implementation and then efficacy testing in line with the strategies.

But again, that's probably a little bit more outside the scope of this presentation.

Today, what I'll briefly run through is a few ideas about how to build a systematic or quantitative investment strategy.

So first things first is we have to come up with a hypothesis of how we're going to invest.

So let's say, you know, somebody sits down and they say, well, I think that the best ASX 200 companies who are moving and leading the market higher may experience brief periods of weakness before continuing their upward march.

Well, we need to look at it and go, okay, what are the quantifiable components of the strategy?

First, we need to identify what the universe is, ASX 200.

Then we need to identify, or how do we identify what a trend is?

We need to define what a period of weakness is, and then reversion, which is before, which is what is written there before continuing the after of March.

So again, that's another part of the strategy.

So we've been able to define it.

So first thing that we need to do is we need to have a universe of data.

We need to have data to do this.

We need the ASX Top 200, including dividends, past and present constituents.

We need to have open and closing prices.

We need to have the volume that traded.

We need any stock adjustments, dividends, splits, consolidations included.

We need it to be accurate so that we can test properly.

You can get this data from many sources.

TIC data, Algo, CSI, Norgate, Refinitiv and Bluebird have all in our experience been some of the best providers.

So second, identification of the trend.

Well, leading the market higher.

How do we statistically and mathematically define what that is?

Well, let's use a simple moving average.

It's the average of the last N days so that we can see that the price is trading higher than that.

A few examples where price is trading greater than the 200 daily moving average, or the prices are greater than the 100 or greater than the 50.

Each of those, one is long, obviously the 200 daily,

all the way down to 50, which is slightly shorter frequency.

So you might have a little bit more noise in your testing process.

But there we go.

By using something like that, we're able to identify what the trend is firstly.

Here's an example of Pro Medicus updated this morning.

You can see the blue line is the daily moving average and you can see that the price is respecting this and is continuing to trend higher.

So the price is greater than the 100 daily moving average in this example, which is what we're going to use.

Second, well, how do we define the period of weakness?

Well, period of weakness, let's use a relative strength indicator.

Again, there's the formula.

Essentially, it is, it compares a security strength on up days versus a security strength on down days.

Again, mathematically, we can test this.

And again, what are we looking for?

Well, we want a combination of all these now.

We want the price greater than the 100 daily moving average.

And then also we want a relative strength indicator less than 30 or less than 20.

Again, in this instance, we're going to use less than 20.

So we want something to be extremely oversold to indicate that we should be looking to buy.

Again, same chart of Pro Medicus.

At the bottom, you can see the relative strength indicator of the last five trading days.

There's been since it has crossed through its 100 daily moving average.

You can see there's been a couple of instances where it's received

an RSI less than 20.

So that was signal, okay we're buying at that level.

Next, okay, well, it needs to revert back to where it was going.

Well, how do we define that the weakness has finished?

Well, again, we need a combination.

We need price is greater than the 100 daily moving average, relative strength indicator was less than 20.

And then let's say we use an indicator where it's come out of it.

And again, using the relative strength indicator, showing a reading greater than 80.

Again, you can see there's instances where this has happened.

There are other considerations which we haven't touched on.

Again, just because of the time that we have available, how much do we allocate?

How do we exit?

So a couple of simple allocation methodologies.

You could do simple equal weight, 20 company portfolio, 5% equal weight.

Volatility based, you know, five times the daily volatility of the position.

Market cap based, liquidity based.

Again, these are all things that you've worked through through the testing process.

The exit methodology again reverts back and I'll jump forward to the next page.

It really jumps back to what we're looking at with our distribution curves.

If we are looking to implement a trend following strategy, we would enter when it has an extreme reading of an RSI of less than 20 and we would look to stay in that investment unless the price closed below the 100 DMA.

that would allow us to have a tail strategy or a right skew.

Big winners but also the potential for some slightly bigger losses as well.

The other version is we could have a mean reversion strategy which is where the RSI is greater than 80 so it rebounds out of that oversold state or you close it where the price is less than the prior day's low.

This would be more of a normal distribution because you're banking profit sooner and your banking loss is very quick as well.

Here is an example of one of the strategies that we've operate within the Frame Longshore Australian Equity Fund, which is quantitative Australian shares.

And you can see the entry signals with a close greater than the 200 daily moving average and a five day RSI less than 20 with obviously a volume parameter because the liquidity, we need to have enough shares to be able to trade it.

By doing and following these signals, you can see that the annualised rate of return works out to be around about 19%, with some strategy performance characteristics here as well.

And you can see, like what I was talking about right at the bottom, you can see the distribution of the curve is very normal, but we have a skew on the right side, which means that we're going to be making money over the long run.

Outside of that, just to summarise, Frankfunds does specialise in quantitative investing.

We essentially offer investment strategies that are reserve institutional investors and we offer these through our two divisions.

Managed accounts,

and managed funds.

And if you would like to find out anything more about it, please reach out at our office or send an email through to admin at framefunds.com.au and we can run through any specifics to do with the strategies, the managed accounts and some of the more complicated things that we go through further on to really build out the robustness of the strategies that we look to implement.

Thanks, Angelina.

Thank you, Hugh.