This trend indicator just issued a "buy" signal in UK markets
The Alternative Investment Market (AIM) has long been the favourite hunting ground of UK private investors. Not only does it host many of the most dynamic young companies in the UK, but the tax incentives (especially the relief on inheritance tax) encourage interest. But it's also probably the most heartbreaking part of the market, as so many companies disappoint.
Perhaps because of this, the index that tracks the aggregate performance - the FTSE AIM All Share Index - is actually an excellent barometer of private investor sentiment. What's more, due to the tendency of sentiment to "trend" I've discovered that the index is a very useful market timing indicator. It's just flashed a "buy" signal for the first time since mid-2021.
The dismal performance of AIM
If you look at the performance of the FTSE AIM All Share since 1996 you'd be astonished. It's declined something like 25% over the last 26 years - see the chart below. Heartbreaker indeed. This rollercoaster performance is not for the faint-hearted.
But it trends...
But if you look more closely, you'll see that this index has a strong tendency to trend. One of the simplest trend following models is to buy when a price chart rises above a moving average, and sell when it goes below. Trend following in this way is prone to failure and false signals, and some indexes give far more false signals than others (for example, the FTSE 100 and FTSE All Share are not for trend followers). However the AIM market can create some forceful signals that persist for long periods of time.
In my investigations, I've found that the optimal moving average to use for trend following AIM is the 120 day (or 6 month) moving average. If you look at this chart, the red line is the moving average and the blue line is the index price. When the blue price line crosses above the red moving average, there is a "buy" signal, and when it falls below, there is a sell signal. As of yesterday, it crossed above the moving average, signalling a buy signal.
Let me whet the appetite for this signal before disappointing you.
If you create a transaction-cost free approach to modelling the profitability of this signal, you find something quite astonishing. It has generated a 4700% return since 1996. In theory that's a 16% annualised return... with low downside volatility. The chart below shows this "cost-less" trend following return - with the index in blue, and the strategy of going long only when the moving average is above the index in orange.
Whenever the above chart is horizontal, the model is "out of the market" - and the signal would have kept a portfolio in cash since late 2021, missing most of the terrible AIM bear market decline.
But...
"If something seems too good to be true, it probably is" Anon
The big problem is that, in the real world, there are trading and tax costs, and as the model creates a bit of whipsaw if you follow buy/sell signals religiously the performance significantly drops.
If using the 120-day moving average, there is an average of 5.5 trades per year (less if using a 6-month average). If you consider the stamp duty on purchases and the typical bid-ask spread, you would be looking at a minimum 2% transaction cost - which would be far higher if you buy less liquid shares - and liquidity is a big problem in bear markets. If you add that 2% cost, then the model returns drop dramatically. If the cost is higher, then it drops further. It can be a lot of work for meagre, but less volatile returns.
Further to this, the model is prone to false signals, especially after declines. It often takes the index a fair while to "base" as sentiment whips up and down. And so buy signals are often followed by sell signals.
Nonetheless, for stock pickers, monitoring a trend / timing model like this can help indicate lower risk market environments for deploying capital. On a personal note, I've been tracking this indicator for the last 12 months, awaiting a signal to do just that. While I'm still feeling quite bearish, I also try to be rational and data-driven. I will be deploying more capital cautiously based on this signal.
About Edward Croft
Co-founder and CEO here at Stockopedia.com. I was a wealth manager, then full-time private investor before setting up Stockopedia. I believe passionately in the power of data-driven investing to improve investment results. Oddly obsessed with the StockRanks.
Disclaimer - This is not financial advice. Our content is intended to be used and must be used for information and education purposes only. Please read our disclaimer and terms and conditions to understand our obligations.
Interactive brokers have a proprietary pattern recognition technology called Trading Central to generate trading ideas. Using AI trading could become more predictable if the crowds are following clear signals. Just a thought.
Soon your car will tell you to stop at the shop for some milk the fridge has alerted.
Thanks Ed - really interesting as always.
in terms of minimising dealing costs though, presumably the point here is that if the indicator relates to the index as a whole, you buy the index and not individual stocks, so is there not an AIM ETF which could be bought, which would have no or a minimal buy-sell spread?
I like this Ed, after all, it's active optimism which moves the market in the right direction. Nobody buying, nothing much going on, it's time for a change.
"is there not an AIM ETF which could be bought, which would have no or a minimal buy-sell spread?"
I don't think there is an AIM ETF available. The nearest ETF to an AIM one might be the iShares MSCI UK Small Cap UCITS ETF (ticker code CUKS) which tracks the MSCI UK Small Cap index. However, some of its holdings (such as Centrica whose market cap is £5.5 billion) are not what I would consider to be a small company.
If we are stockpickers it might be useful to look at sector momentum as a guide to which sectors we should overweight in our portfolios to take advantage of the recovery. The chartlets below show average sector momentun rank for the companies that I follow. (Right click/open in new tab to see larger charts)
Average momentum rank is falling in Energy and Utilities and remains low in Telecoms. Consumer Defensives and Financials show no real trend. There is an upward trend in Basic Materials, Industrials, Technology and Healthcare but the one I find most striking is Consumer Cyclicals, where I will be concentrating my search for opportunities.
How are you investing in this Edward? I cant find any index trackers of AXX and can’t believe you’d buy the 731 constituents of the index.
You can't invest in it. There's no tracker for the AIM All Share (I don't think). It's the signal in general I'm interested in, especially as a barometer of sentiment and general market direction.
Hi Ed,
What might be a nice idea would be for Stocko to publish when the AIM All Share 120 DMA trend indicator produces a "Sell" signal for the UK. That way investors can be alerted to the change in the trend and then decide what they want to do with their investments.
David.
I did have a look at the Aim index. It is useful feature on Stockepedia that you can select the index then put in order by market cap with the largest at the top, then select the mini charts. Here are the charts : https://www.stockopedia.com/in...
If you change it to 6 months it is easier to see what is happening in the last 4 months.
There are fairly large rises in the big mcaps but the picture overall in rather mixed. The Aim index is a weighted index.
There have also been a number of takeover bids, I think I saw 5 on going.
I also looked at some sectors and CC which has also risen there seems to be a more consistent bounce in share prices from around mid October.
I'm not convinced by this bounce and have been selling more than buying recently, but I'm not good at calling the markets.
"In my investigations, I've found that the optimal moving average to use for trend following AIM is the 120 day (or 6 month) moving average." So that was the best fit to historical data. In 3 years time might you not find that the best fit was instead a 60 day moving average? Just a thought, I am no expert but, as you say, taking into account real-world dealing costs you have to be very sure of your model. AIM is so sentiment-driven and some director guidance is so overly optimistic as to verge on the fraudulent. The best thing with AIM in my opinion is to get to know the company and talk to the directors. Make sure they have skin in the game and are not overly rewarding themselves.
@KhrisCarter - the test is over 26 years, so I don't think 3 years of data will make too much difference - but you are right that the nature of a market can change.
I don't though believe that the type of AIM Market listings, nor the individual investor interest in it, will vary much unless there's a huge regulatory and tax incentives shift. The market behaviour will be quite stable until there's a significant environmental shift.
Ed thank you for a very interesting article. Please can you elaborate on the following questions? I would love to know a) How do you determine whether any index has a strong tendency to trend or not? b) Which indices were examined for tendency to trend? c) How you determine that the optimal moving average to use for trend following AIM is the 120 day (or 6 month) moving average? d) Which other moving averages were evaluated for trend following capabilities?
@WarrantStar - I have a pretty simple measure for whether an index trends or not.
Does a simple moving average crossover, with a sensible lookback period (100-200 days), yield profits significantly above the index without too much choppiness. (Not very scientific).
I can't pretend I did the most extensive study. I checked the FTSE 100, the FTSE All Share, the FTSE AIM All Share and the FTSE 250. The AIM All Share is by far the best signal generator - albeit with the caveat that it would be the most expensive to trade in or out of.
I've got a script so I can run these studies pretty quickly. I've just run the FTSE Small Cap ex-IT index... and it's pretty good after transaction costs. The median bid-ask spread on that index is only 50bps... so if I model a 1% transaction cost on every trade (including stamp duty) it still generates well above the index returns on a 120 day moving average crossover model.
There are a few nasty drawdowns where the index gets choppy - look at 2009-2012 where the model lost 28% while the market went sideways. It's quite hard to keep the faith in moments like that!
*Past performance is no indicator of future performance. Performance returns are based on hypothetical scenarios and do not represent an actual investment.
This site cannot substitute for professional investment advice or independent factual verification. To use Stockopedia, you must accept our Terms of Use, Privacy and Disclaimer & FSG. All services are provided by Stockopedia Ltd, United Kingdom (company number 06367267). For Australian users: Stockopedia Ltd, ABN 39 757 874 670 is a Corporate Authorised Representative of Daylight Financial Group Pty Ltd ABN 77 633 984 773, AFSL 521404.
just bought some this week