Hurco Companies, Inc. released a 10-Q report today which covers the company’s fiscal second quarter ending on April 30, 2009. I covered the initial release of Hurco’s earnings for the quarter in an article last week. In that article, I mentioned some concerns about an unexplained drop in Hurco’s warranty provision in Q1 and the need to closely examine the warranty provisions once the 10-Q for the second quarter is released. Now that we have the 10-Q, let’s take a look at the warranty provision concern.
Calculation of Warranty Provisions
The company does not provide much in the way of details regarding how warranty provisions are estimated for any given period. In the latest 10-Q, the following explanation is provided:
We provide warranties on our products with respect to defects in material and workmanship. The terms of these warranties are generally one year for machine labor and service parts. We recognize a reserve with respect to this obligation at the time of product sale, with subsequent warranty claims recorded against the reserve. The amount of the warranty reserve is determined based on historical trend experience and any known warranty issues that could cause future warranty costs to differ from historical experience. The warranty reserve may vary due to changes in sales volume, product mix and sales by region.
The final line in the explanation appears to be an addition compared to the language used in the 10-Q for the first quarter. I contacted Hurco’s management and received a response stating that they would consider adding more detail. Unfortunately, the additional detail merely states the obvious and does not add much that is useful to the analyst, particularly in light of the apparent change in management’s methodology for recent quarters.
Warranty Provision History
Let’s examine the history of Hurco’s practices when it comes to warranty reserves and provisions over the past five years. The exhibit below shows data for the last five full years as well as the first two quarters of the current fiscal year:
In my opinion, the appropriate metric to use when evaluating warranty provisions for a particular period is the ratio of the provision to cost of goods sold. The alternative would be to take the ratio of the provision to sales. However, since the company’s cost of sales is a more accurate reflection of what it would cost to provide warranty services and parts, I have selected that metric for this analysis.
For the five full fiscal years from 2004 to 2008, warranty provisions as a percentage of cost of sales ranged from 2.06% to 3.23%, with a drop between 2004 and 2005 followed by relative stability in the low 2% range over the past four years. During the five year period, aggregate provisions totaled $11,488,000 and warranty charges totaled $10,013,000. Based on historical precedent, it would appear that management reserved properly for warranty provisions during this time frame.
In the first fiscal quarter of 2009, warranty provisions as a percentage of cost of sales dropped to 0.29%. For the latest quarter, the measure came in at 1.25%. For the first half, warranty provisions as a percentage of cost of sales measured 0.71%. This appears to be a precipitous drop from the reserves taken in prior periods. Either a change in methodology has been adopted or management believes that the product mix or regional sales mix will materially impact warranty charges in the future.
If one takes the average warranty provision as a percentage of cost of sales for the last five full years (2.34%) and applies this number to cost of sales in Q1 and Q2 2009, we can infer that warranty provisions should have been far higher than recorded. In aggregate, there would appear to be a shortfall of $573K for the first half of fiscal 2009.
What Does This “Prove”?
This analysis, taken in isolation, does not in any way “prove” that earnings are being managed or that the warranty provision is not correct. Management could well have solid reasons for the change in approach between prior years and the first half of the current year. However, we are not told of any such reasons for such a dramatic change which leaves the suspicion that earnings are being managed during a particularly rough period for Hurco from a business perspective.
In the absence of convincing reasoning to the contrary, I will assume that warranty provisions should be adjusted to reflect prior historical averages and assume that earnings were actually smaller than reported in the financial statements (or more accurately, net loss for the first half should be higher than reported). My view is that a change in methodology of this magnitude requires additional disclosure if we are to accept the figures in the financial statements without making appropriate adjustments for historical experience.
We have liquidated our position in Hurco this morning at an average price of $14.75. The decision was not made based on the warranty provision issue alone. There are two additional primary factors that have changed the equation enough to warrant a sale of the position:
First, as I mentioned in the article last week, I have growing concerns about the fact that management has not trimmed inventory aggressively to reflect the precipitous drop in sales so far this year. The probability that inventory will be written down has increased significantly. If this occurs, the drop will be reflected in Hurco’s current assets and obviously would have an impact on metrics such as tangible book value per share. The discount to Hurco’s reported tangible book value per share was one of the primary factors leading to the initial purchase decision.
Second, I am growing less comfortable with Hurco’s regional sales mix based on the condition of Europe’s overall economy. Hurco is heavily exposed to Europe with over 63% of sales coming from that region in the first half of the current fiscal year. Europe has generally been less aggressive in terms of monetary and fiscal stimulus compared to the United States and this will have an impact on economic growth going forward. While I never make investment decisions based on macroeconomic factors alone, Hurco’s exposure to Europe is a contributing factor in the decision.
It is possible, and perhaps even likely, that Hurco will enjoy a rebound in business and share price over the next several years. However, my initial purchase rationale for the position was based on the company’s apparent discount to net current assets and tangible book value. With my suspicion that the company’s balance sheet is now more impaired that the reported numbers show, I can either exit the position or change my rationale for holding the shares from a net current asset play to a call on the eventual growth in the global economy in general and Europe in particular. In my experience, changing the investment rationale for a position is nearly always a bad idea in the long run.
The position in Hurco advanced 11.07% from the purchase on April 7 while the S&P 500 advanced around 14% in the same period. I normally do not consider investments in special situations (such as net current asset plays) unless I believe that there is at least a 50% upside, so the result is disappointing from that perspective. In addition, underperforming the S&P 500 is always disappointing. Nevertheless, all things considered, there are worse things than exiting a position with gains that probably should not have been acquired in the first place had a more thorough review of all factors been considered initially. Should Hurco’s stock price return to the lows of March, it may well be worth looking at again even considering the potential concerns brought up here; however, at the current quotation, the margin of safety is not sufficient.