XML 22 R10.htm IDEA: XBRL DOCUMENT v3.20.2
RECEIVABLES
9 Months Ended
Sep. 30, 2020
Receivables [Abstract]  
RECEIVABLES RECEIVABLES
Accounts receivable consisted of the following (in thousands): 
September 30, 2020December 31, 2019
 (unaudited) 
Trade accounts receivable$176,583 $192,743 
Unbilled receivables47,411 62,864 
Allowance for credit losses(9,633)(9,990)
Total$214,361 $245,617 
ASC 326 applies to financial assets measured at amortized cost, including trade and unbilled accounts receivable, and requires immediate recognition of lifetime expected credit losses. Significant factors that affect the expected collectability of our receivables include macroeconomic trends and forecasts in the oil and gas, refining, power, and petrochemical markets and changes in our results of operations and forecasts. For unbilled receivables, we consider them as short-term in nature as they are normally converted to trade receivables within 90 days, thus future changes in economic conditions will not have a significant effect on the credit loss estimate. We have identified the following factors that primarily impact the collectability of our receivables and therefore determine the pools utilized to calculate expected credit losses: (i) the aging of the receivable, (ii) any identification of known collectability concerns with specific receivables and (iii) variances in economic risk characteristics across geographic regions.
For trade receivables, customers typically are provided with payment due date terms of 30 days upon issuance of an invoice. We have tracked historical loss information for our trade receivables and compiled historical credit loss percentages for different aging categories. We believe that the historical loss information we have compiled is a reasonable basis on which to determine expected credit losses for trade receivables because the composition of the trade receivables is consistent with that used in developing the historical credit-loss percentages as typically our customers and payment terms do not change significantly. Generally, a longer outstanding receivable equates to a higher percentage of the outstanding balance as current expected credit losses. We update the historical loss information for current conditions and reasonable and supportable forecasts that affect the expected collectability of the trade receivable using a loss-rate approach. We have not seen a negative trend in the current economic environment that significantly impacts our historical credit-loss percentages; however, we will continue to monitor for changes that would indicate the historical loss information is no longer a reasonable basis for the determination of our expected credit losses. Our forecasted loss rates inherently incorporate expected macroeconomic trends. A loss-rate method for estimating expected credit losses on a pooled basis is applied for each aging category for receivables that continue to exhibit similar risk characteristics.
To measure expected credit losses for individual receivables with specific collectability risk, we identify specific factors based on customer-specific facts and circumstances that are unique to each customer. Customer accounts with different risk characteristics are separately identified and a specific reserve is determined for these accounts based on the assessed credit risk.
We have also identified the following geographic regions in which to distinguish our trade receivables: the (i) United States, (ii) Canada, (iii) the European Union, (iv) the United Kingdom, and (v) other countries. These geographic regions are considered appropriate as they each operate in different economic environments with different foreign currencies, and therefore share similar economic risk characteristics. For each geographic region we evaluate the historical loss information and determine credit-loss percentages to apply to each aging category and individual receivable with specific risk characteristics. We estimate future expected credit losses based on forecasted changes in gross domestic product and oil demand for each region.
We consider one year from the financial statement reporting date as representing a reasonable forecast period as this period aligns with the expected collectability of our trade receivables. Financial distress experienced by our customers could have an adverse impact on us in the event our customers are unable to remit payment for the products or services we provide or otherwise fulfill their obligations to us. In determining the current expected credit losses, we review macroeconomic conditions, market specific conditions, and internal forecasts to identify potential changes in our assessment.
The following table shows a rollforward of the allowance for credit losses:
 September 30, 2020
 (unaudited)
Balance at beginning of period1
$(11,400)
Provision for expected credit losses(1,160)
Write-offs2,841 
Foreign exchange effects86 
Balance at end of period$(9,633)
_________________
1 Includes $1.4 million due to the initial adoption of ASC 326 as of January 1, 2020.