Headlines that include numbers garner too much attention and not enough skepticism from investors. The EPS (earnings per share) number headline is a classic example: when EPS is even a penny more than expected, stocks often move higher.

Loose accounting rules make it child’s play to manufacture a few cents in earnings, and company executives are well-practiced at managing analysts’ expectations ahead of time. But it hasn’t stopped the emergence of an entire industry in whisper numbers, including sell-side vs. buy-side EPS estimates.

Even with the shortcomings of EPS figures more widely understood these days, investors tend to blindly believe a host of other headlines with numbers. The biggest risk is when companies manufacture their own sets of books and figures, using them to supplant standardized metrics when reporting to investors. Remember Nortel? At the same time its press releases showed billions of dollars in “net earnings from operations,” its audited financial statements listed billions in net losses.

While Nortel vanished, its legacy did not. The problem continues in modern incarnations of the same trick. Recent columns have focused on the risks of believing in adjusted EBITDA, and just last month, free cash flows (see “Problems with free cash flows,” Advisor’s Edge Report, March 2014).

The list goes on

Industries also tend to have their own headlines with numbers. In the railroad sector, a lot of attention has been given to the operating ratio (OR), which is calculated by expressing operating expenses as a percentage of revenue—the lower the better.

OR ignores the capital side of the business, and is more focused on management efficiency. It requires assumptions to split ongoing operating costs from total costs. While there is general agreement among the Class I North American railroads as to which costs are excluded, there can be some differences.

When one-time costs are excluded to report an adjusted operating ratio, they are generally labeled as such, minimizing the potential to mislead investors. Adjusted items might include legal settlements, acquisition expenses, derailment costs, asset impairments, and restructuring.

Among Class I North American railroads, adjusted ORs fluctuate between 60% and 75% on a quarterly basis, depending upon seasonality, efficiency of the operator, external events such as extreme weather, and management accounting assumptions (see “2013 adjusted operating ratios,” this page).

Company executives will estimate dozens, if not hundreds, of key accounting assumptions that naturally have an impact on reported results.

For instance, depreciation rates on different types of assets will impact ORs. However, most of these assumptions are irrelevant to investors.

When examining the impact of accounting assumptions on reported figures, it’s critical to focus on gaps in assumptions between peers, or shifts in assumptions from period to period within the same company.

Gaps between peers are relevant because a more aggressive assumption at one company can make it appear better than its competitors. Meanwhile, shifts in assumptions can move the needle in terms of year-over-year comparisons, which tend to impact investor perceptions materially. Accounting assumptions related to company pension plans can have a significant impact on reported results.

Key management estimates include the interest rate and expected return on pension plan assets, both of which are used to determine pension expense, which impacts OR.

Canadian Pacific’s accounting

Under new management since 2012, Canadian Pacific (CP) has been successful in driving down its OR from 81.3% in 2011 to 69.9% in 2013 through a combination of items, including labour attrition, increased train lengths and longer sidings. As a result, the stock price rose more than 150% during that time.

This year, CP wants to lower its OR to 65% at most, and it already seems to be halfway there, thanks to accounting. Instead of reporting expenses for its pension plan in 2014, management expects to report pension income (or negative expenses).

While it seems counterintuitive that having a pension plan can result in negative expenses, that’s the way accounting works, so it should be watched closely at all companies.

For instance, when calculating pension expenses, management will deduct what it estimates the company will gain as a return on pension plan assets. Making pension income is a matter of estimating more market gains on assets than there are expenses from operating the plan.

The company can accomplish this in two ways.

  1. A higher expected rate of return on plan assets could produce enough unrealized gains to simply offset the accrued employment costs.
  2. Likewise, a change in the discount rate also impacts reported expenses, but in a much more complex way. The discount rate affects service costs and interest costs in opposite directions, separately impacts the obligation on which interest costs are based, and also potentially triggers actuarial gains and losses. Taken together, an increase in interest rates usually results in lower pension expenses.

While there are many moving parts, the upshot is that if CP were to have recorded pension income of $50 million in 2013 (the same amount that it expects for 2014), we estimate its operating ratio would have dropped from 69.9% to 67.6% (which is already half the improvement management is aiming for this year).

Even though it has nothing to do with efficiency, and everything to do with accounting, a 230 basis point improvement in OR is an impressive result. And it’s bound to make headlines everywhere this year.

2013 adjusted operating ratios for North American Class I Railroads

Canadian National Railway68.4%60.9%59.8%64.8%
Canadian Pacific Railway75.8%71.9%65.9%65.9%
CSX Corp.70.4%68.6%71.5%73.2%
Kansas City Southern70.5%69.0%67.8%68.1%
Norfolk Southern74.8%70.2%69.9%69.4%
Union Pacific Corp.69.1%65.7%64.8%65.0%