ArbMaker Premium Newsletter issue #12.
View this email in your browser
Healthsouth (HLS) and Cooper Companies (COO) are healthcare firms facing the new Trump context of repeal or replace the Affordable Care Act, popularly known as Obamacare.

It is HLS that has the real exposure to the politics of this but it is noticeable, after an initial drop of over 10% in its stock price, that it is back to its 7 November level. Rivals who have benefited more from the expanded coverage the Act has brought to some states have been less fortunate (like Tenet Healthcare).

COO is principally a specialist in eye care but with an important surgically-delivered range of products for women's health care. These are delivered via hospitals, fertility clinics and doctors' offices.

As the fundamental data currently stands neither firm puts the other in the shade. However, the fall-out on healthcare policy in the US is a larger risk for HLS than COO.


HLS uses the word 'rehabilitation' a lot as a descriptor of its hospital services. It is apt in a larger sense: under the soon to be retiring CEO Jay Grinney HLS came back from a massive accounting fraud in the early 2000s under then CEO Richard Scrushy. After the inevitable earnings restatements it has been digging itself steadily (and impressively) out of a retained earnings deficit ever since.

Today its metrics look decent. It has more leverage than COO but is a more productive beast with it. That it also looks significantly cheaper than COO (using EV/EBITDA) therefore may seem surprising. But valuations of US hospital stocks have been falling all year relative to the broader market as the election - and fate of Obamacare - drew nearer and nearer.

For its part, COO reports on 8 December with a very encouraging set of quarterly earnings reports already behind it in 2016. It may be less productive than HLS but its improvement on that count has been sharp. On current trends it is a tough short in front of earnings. That said, its price has been drifting downwards since hitting a record high on the back of its last 10-Q.

Today the pair's spread is near a short HLS / long COO signal. HLS has the shadow of uncertain health policy upon it; and COO the burden of high expectations. Subscribers may recall that this was also the case when we traded it last September. On that occasion both legs won for a total gain of 6.6% under standard Interactive Brokers margin. The CFD equivalent trade made 14.5%.



This is an end-of-day analysis. We are using a strategy that incorporates a directional check and a double criteria entry.

If you are running ArbMaker software the pair can be tracked using the file in the Execution section below. The Healthsouth investor relations page is here. That for Cooper is here.

Historical Performance

On the defined set-up HLS/COO has traded twelve times since January 2013. Ten trades won. Total average return per trade (including losers) was 1.0%. That calculation does not account for margin traders may have access to.

Trade duration averaged 13 calendar days. The worst drawdown recorded was 4.1%. Average drawdown was 1.7%.

Selected performance data, HLS vs COO

Click to enlarge. Accounting data comes from SEC Edgar filings, updated in near real time. Balance sheet data uses the latest available quarter; other data is shown on a trailing twelve month (TTM) basis. Some metrics are proprietary but based on straightforward concepts.


To follow the evolution of this pair download and import:
  • This single item portfolio called Issue_12_selection.txt containing HLS/COO as we have structured it. This uses $6,500 as the default amount for Y but the size is user-definable.
The strategy set-up is integrated with this file.
The file uploads via the Utilities->Import Data button. Remember to click Refresh in the Portfolio screen to update.
This configuration uses a time-varying beta regime and is therefore money neutral: every $100 in one requires $100 in the other to balance the hedge. We run our tests under IQFeed data.
Both symbols are available in CFD form with Interactive Brokers (for eligible accounts).


Current status

HLS/COO is currently on WATCH status. We'll issue a flash alert on any triggered signals.

Research review

The Beneish M-score, now about 17 years old, is well known in accounting and fundamental investing circles as a method of assessing the likelihood of financial manipulation. It came to attention mainly due to its having flagged Enron as a fraud in waiting. Had it been around in the mid 1990s when Healthsouth were being naughty capitalizing expense items it would have flagged them too.

It’s simple to calculate from published financial statements and some financial websites even do it for you. It is a metric that lends itself, especially in combination with other data, to the long/short trading process.

Although the main goal of this "To catch a thief" paper was to test the M-score in an out-of-sample analysis (which indeed validates Beneish’s model) it also lays out a few ways how such combinations can enhance the potency of the ratio.

In issue #5 of this newsletter we covered one such complementary ratio, percentage accruals. The sections of this paper covering the relationship between it and the M-score are particularly valuable to long/short traders.

Of course, context matters greatly and establishing it can take some digging. But in an era where the accounting data is free and readily available from the Securities and Exchange Commission’s website that is a small cost to bear.

Trading update

The deterioration in performance since earnings season and through the November 8 US Presidential result means this newsletter now trails the S&P 500 by 300 basis points since its 10 June inception.

Put another way, the letter is marginally better than flat and the S&P 500’s entire 3.1% gain has come since 4 November. Compared to long/short and relative value indices through October via sites like Eurekahedge the story is similar (but somewhat worse) to ours.

In summary, the Tide of Trump has raised a great many boats to the detriment of stock picking and - depending on who’s talking – has either (i) created pockets of temporary price turbulence or (ii) heralds economic regime change.

The verdict will sit, probably, between those two opinions.

Flash Alerts - a guide

We are leaving this section in our newsletters for reference purposes

Our flash alerts reflect our trading positions and are highly selective. When we send a flash alert it means several trading filters have been met - this is especially true for entering trades.

The first filter relates to the stat profile of the pair in question and its assigned trading strategy. This filter comes from the ArbMaker software and – this is important – it does not mean we will simply take the trade.

Thus if you are using the tracking files we provide you may see the signal at the same time. But it does not mean we are taking the position. Only a Flash Alert notification confirms that.

What are the additional filters we look at once ArbMaker signals? Some are proprietary and based on accounting fundamentals; some apply only to intraday trades; and some look carefully at news flow and sentiment. If these all line up we will take a position and send a flash alert.

Exit strategy (stops or profits) will usually follow the ArbMaker signal but there are exceptions. These usually relate to either a higher than expected rate of mean reversion or spread behaviour clearly inconsistent with history. In either case we may close the position and send a flash alert.

And here's what they look like

Delivery of Flash Alerts

Alerts go by email and via a private Telegram app channel which we are alpha testing. If you wish to get the alerts via Telegram please email us and we will send you an invite to the channel.
Publication schedule update
Our next issue #13 will be released on Tuesday, 6 December, 2016.
Feedback? Email us.
Copyright © 2016 ArbMaker, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list

Email Marketing Powered by Mailchimp