Is the Stockopedia Community influencing the market?

A couple of weeks ago I received the following questions from one of our long term subscribers: “Now that Stockopedia is so successful, do you feel that you are influencing the market as a whole?”  “Does the availability of high-grade information to private investors cause them to all move together under your guidance?“ “Are you increasing the volatility of the market?”

I hear these kinds of questions time and again, and until now I’d mostly disregarded them. We’ve always punched above our weight, but as a niche player our market impact was intuitively low.

But over the last 12 months our subscriber base has grown by more than a third, the readership of the blogs has grown to all time highs and the StockRanks have continued to perform stunningly well.

Given these facts, and the fact that I use the StockRanks myself as my core investment process, I feel some responsibility to answer these questions.

The Study

Part of the Stockopedia service is a suite of portfolio tracking tools. I would say that less than half of subscribers use them but they serve as a good proxy for the proportional weight of money moving into certain shares.

To help answer these questions for the community, I’ve aggregated the last 3 months of buy and sell transaction data from the portfolio systems.

I’ve limited the search to London Stock Exchange securities. While do have a growing army of international investors on the site, from Canada to Oz, as a UK founded site, the majority here are British.

I’ve also excluded some portfolios that are clearly marginal or erroneous. We do have quite a number of fund managers using the service, and their portfolios are normally £200m+ in value which skew the results. I also excluded all the outlier transactions which could have been entered by fat fingers. While the resulting data set is not perfect, it’s good enough to answer quite a few questions.

The Results

1/ Breadth

The first surprising result is quite how many different shares are traded by the community. If you exclude all collective investment vehicles, dual lists and other odd listings there are 1472 common shares on LSE. The community traded 1294 of them. That’s 88% of the market’s shares in a single quarter.

2/ Concentration

Of course there are hundreds of these shares that are barely traded at all. The chart below shows this clearly. The horizontal axis shows the number of trades logged in the quarter, while the vertical axis shows the number of different stocks traded that much. The large vertical bar on the left shows that more than half the stocks traded had less than 15 trades logged during the quarter. But there were also some names that were traded hugely.

There are though 2 ways to consider the trading interest in a stock. The first is the number of trades, and the second is the value of trades. The most traded shares in both cases differ quite substantially and we’ll look at each in turn.

2a/ Most Traded by Trade Count

Firstly, the top ten most traded shares, by number of trades during the quarter were:

In the above graphic I’ve included whether the Small Cap Report has been covering the stock at all favourably. Paul & Graham who write the report never tip shares, but they are often opinionated in their analysis and their own positions. We can see that about half of the top ten traded stocks on the site by trade volume have had some coverage by Paul or Graham during the quarter. It is also worth noting that a few of the other shares (Morgan Sindall and XLM Media) not covered by the Small Cap Report are also current holdings in Roland Head’s Stock in Focus Model portfolio.

2b/ Most Traded By Value

Of course there’s an army of small shareholders out there who trade in very small size. If you actually multiply the number of shares traded by the share price you can estimate the value of the trades during the quarter. If we do so, a very different picture emerges:

Most of these shares are large cap blue chip names, and there’s very little editorial coverage on these names, if at all on the Stockopedia site.

It’s great to see how independent our community is in making their decisions. This is very much a DIY/DYOR site and we do not ever make stock recommendations. We provide the subscription tools and data to empower everyone to make their own independent decisions. The community appears to be informed by the editorial, but not overly influenced by it. There are a lot of independent decisions going on.

Influence of the StockRanks

The following chart shows the average StockRank of stocks traded during the quarter - buys in blue, sells in red. Given the huge breadth of trades made during the quarter (88% of the market traded), it should be little surprise that the average StockRank of stocks traded is around the 54 mark. But if we ‘value weight’ the trades to show the weight of money a different picture emerges. There’s a much larger weight of money going into higher StockRank shares. The average StockRank of buys is 71, while the average StockRank of the top 100 most bought stocks is only slightly higher at 76.

Given that the 90+ StockRank set of shares has most outperformed the market, It may surprise some that the average StockRank purchased is so low. . But let’s look a bit deeper. If we take the top 100 shares traded by value and plot them as a histogram we find the following:

About 36% of the top 100 traded stocks had a StockRank greater than 90. It’s a sizeable proportion, but it’s clearly not excessive. I believe these insights may help dispel the notion that there is herding amongst high StockRank names. This community is much more diverse than many believe.

We can also view the transactional data using the StockRank “Style” feature we’ve recently released. Only 11% of the money invested by Stockopedia Subscribers has gone into “Losing Style” shares, with 63% into “Winning Styles”. I take my hat off to the community here - we’re all managing to avoid the dregs and skew our portfolios to higher probability investments.

The Market Impact of the most traded stocks

I’m now going to tread into the area of ‘market impact’. This is tricky to assess for a few reasons. Firstly, not everybody subscribing to (or reading) Stockopedia is tracking their portfolio in the system. Many people also use spread betting or CFD accounts which we currently don’t provide for. So any data I publish here really needs to have a multiplier added. What that multiplier should be is very hard to assess, so I will only publish the facts that I know.

I’ve defined ‘market impact’ as the value of Stockopedia Community trades divided by the total value of trades in the market. So if the Stockopedia community has traded 100 of every 1000 shares traded during the quarter, the ‘market impact’ is 10%.

From the above chart we can see that, aside from a few outliers, the community market impact is fairly small. Less than 5% of volume for all but 4 stocks. The average market impact of the top 100 is 1.14%.

Even if we conservatively presumed that only 1 in 5 trades by the Stockopedia community are logged in the portfolio tracker, the market impact would still average less than 6% of volume. Even across all traded stocks (not just the top 100), there are only 11 shares where the market impact has been greater than 10% and these tend to be in illiquid micro-cap shares.

Market Impact on the top 100 by StockRank

It’s so easy to find the highest ranking shares on Stockopedia that the assumption must be everyone is using them. After all, you just have to click the “ranks” tab to find the top 100 ranked shares. The historical performance is published for all to see in the tab next to it and we blog about the philosophy behind them constantly. Surely everyone is trading these?

Well if we flip our market impact analysis and start with the top 100 ranked shares, here’s what we find:

There’s a slightly higher market impact across these stocks but barely. On average 1.79% of total volume (median 0.48%) has been traded by the Stockopedia community. This is hardly excessive, in fact in most cases it’s barely noticeable.

To the left of this chart, IEnergiser is the anomaly with nearly 30% of the volume traded, but this is a highly illiquid, tightly held stock. While 2 micro-caps Titon Group and North Midland Construction both have had more than 10% of their total volume traded by the community. That’s only 3 out of 100 top ranked shares seeing high participation rates by the community.

What’s more, the levels of investment into these names is not linear. Of the £50m invested in the top 100 StockRank names over the quarter, 50% was invested into only 14 of those names. A third of the top 100 by StockRank took in less than 2% of the total.

This is clearly illustrated by this chart.

Conclusions

It’s been fascinating for me to run these numbers. I’ve often wondered at what point the size of the UK subscriber numbers would become a problem, but it doesn’t seem like we’re anywhere near it. The market impact is low, and there’s room for many multiples of the current subscriber numbers unless trading behaviour changes considerably.

Our community is mostly full of DIY self-directed investors who all use the tools in different ways. Quantitative investors like me may be more likely to systematically purchase shares purely due to high StockRanks. Qualitatively driven investors are more likely to take other softer factors into account - such as business trends, management and sentiment. My last poll suggested that 80% of subscribers classify themselves as qualitatively driven investors. While my own mission is to grow the 20%, I don’t think we’ll see an enormous behavioural shift for years to come.

The prevalence of qualitatively driven investors in the community may explain the slight aversion to higher ranking shares. High StockRank shares are often psychologically hard to buy. The StockRank selects for Quality, Value and Momentum. But Quality shares are often rather dull and unexciting, Value shares often have perceived problems, while Momentum shares are often at new all time high share prices - nobody likes buying these kinds of stocks. I’m sure this is a reason why there isn’t nearly as much buying amongst high StockRank shares as we might expect.

I hope you’ve found these insights useful. I’d be fascinated to hear your comments.