One big miss, and an accounting brouhaha

It was American baseball legend Satchel Paige who said "Don't look back - something might be gaining on you." I like to look back on recent Vox columns to keep myself honest, but some recent developments have already given me indigestion. In some cases, shareholders didn't get nearly as upset with companies as I thought they should. And there's been at least one terribly untimely stock recommendation.

First, the worst call: I knew there was a place for a Big Lots-style retailer in the Canadian market, and that prompted me to take a favourable look at Liquidation World in April when it was trading at 50 cents a share. Turns out the place is for Big Lots itself, which is buying Liquidation World … at 6 cents a share.

The retailer's first earnings report after my column was a disaster, with an admission that it had breached its loan agreements and couldn't afford to restock its stores. At least I warned readers that it was possible to lose their entire investment in Liquidation World - that's the only yardstick by which a loss of 88 per cent in 40 days is defensible. Sneak preview: Look for this one on my year-end list of mistakes.

Earlier in the year, I criticized pay plans at two Alberta energy companies, TransAlta and Encana. At TransAlta, I objected to a performance share plan that allowed executives to keep a portion of their stock award even when the company fell between the 25th and 50th percentile in total shareholder return among its peer group of companies. At Encana, the company created a new performance-share plan, but in doing so, removed the performance criteria from its stock-option awards. I suggested shareholders of the company should send a message in the two companies' upcoming say-on-pay votes.

They did, and the message was that they don't agree with me: Encana got 91 per cent approval for its compensation philosophy, while TransAlta yielded an 86-per-cent favourable vote. (An earlier column criticizing bonus pay at Suncor was similarly disregarded, as shareholders gave the company 94-per-cent approval.)

Similarly, I took issue with a pay plan at Newport Partners Income Fund, a former income trust that had dropped from a $10 offering price to about 50 cents. The fund issued options equal to about 20 per cent of the company in just 14 months and was seeking shareholder approval for a plan to issue the second batch of options.

It got shareholders' blessing, but not overwhelmingly: Just 72 per cent voted yes. If you take out the votes of the insiders, the company said, the vote was just 64 per cent in favour.

Now here's an update about which I'm less chagrined. In March, I wrote about SNC-Lavalin Inc.'s accounting for transactions between its engineering and construction ("E&C") business and what will soon be a wholly-owned subsidiary, Alberta power company AltaLink.

SNC-Lavalin had said it intended to continue an allowed practice under Canadian GAAP to book "intra-company" profits with a regulated subsidiary as it adopts International Financial Reporting Standards, which make no such explicit allowance. This, says Veritas Investment Research analyst Michael Yerashotis, threatens to confuse investors trying to value SNC-Lavalin based on its reported earnings.

The column prompted a storm in Canada's forensic accounting community, when Al and Mark Rosen's Accountability Research Corp. said in a report that such concerns are "overblown," with "no quantifiable impact on valuation." The Rosens believe that if SNC-Lavalin's E&C business is charging AltaLink fair market prices for its services, then no profits need be eliminated, because they are no different from those received from a third party.

Mr. Yerashotis respectfully disagrees in a follow-up report, saying there is a significant difference in the timing of cash flow between the revenue from AltaLink and from third parties. If SNC-Lavalin - hypothetically - charges AltaLink $100-million for a project that costs them $82-million, says Mr. Yerashotis, that $18-million in pretax profit should be considered more like a cost savings that should be accounted for over the up-to-55-year life of AltaLink transmission assets.

The difference in that one hypothetical project, from a valuation perspective with a 15 price-to-earnings ratio? More than $200-million per SNC-Lavalin's choice of accounting, compared with about $9-million from the Veritas approach. I'm with Mr. Yerashotis on this one.