5. Goodwill

The changes in the carrying amount of goodwill for our subscription and software reporting unit during the nine months ended March 31, 2016 and fiscal year ended June 30, 2015 were as follows:

	Amount
	(Dollars in Thousands)
Balance as of June 30, 2014:
Goodwill	$	84,845
Accumulated impairment losses	(65,569		)

	$	19,276


Effect of currency translation	(1,916		)


Balance as of June 30, 2015:
Goodwill	$	82,929
Accumulated impairment losses	(65,569		)

	$	17,360


Effect of currency translation	(579		)


Balance as of March 31, 2016:
Goodwill	$	82,350
Accumulated impairment losses	(65,569		)

	$	16,781

We test goodwill for impairment annually (or more often if impairment indicators arise), at the reporting unit level. We first assess qualitative factors to determine whether the existence of events or circumstances indicates that it is more likely than not that the fair value of a reporting unit is less than its carrying amount. If we determine based on this assessment that it is more likely than not that the fair value of a reporting unit is less than its carrying amount, we perform the two-step goodwill impairment test. The first step requires us to determine the fair value of the reporting unit and compare it to the carrying amount, including goodwill, of such reporting unit. If the fair value exceeds the carrying amount, no impairment loss is recognized. However, if the carrying amount of the reporting unit exceeds its fair value, the goodwill of the unit may be impaired. The amount of impairment, if any, is measured based upon the implied fair value of goodwill at the valuation date.

Fair value of a reporting unit is determined using a combined weighted average of a market-based approach (utilizing fair value multiples of comparable publicly traded companies) and an income-based approach (utilizing discounted projected cash flows). In applying the income-based approach, we would be required to make assumptions about the amount and timing of future expected cash flows, growth rates and appropriate discount rates. The amount and timing of future cash flows would be based on our most recent long-term financial projections. The discount rate we would utilize would be determined using estimates of market participant risk-adjusted weighted-average costs of capital and reflect the risks associated with achieving future cash flows.

We have elected December 31st as the annual impairment assessment date and perform additional impairment tests if triggering events occur. We performed our annual impairment test for the subscription and software reporting unit as of December 31, 2015 and, based upon the results of our qualitative assessment, determined that it was not likely that its fair value was less than its carrying amount. As such, we did not perform the two-step goodwill impairment test and did not recognize impairment losses as a result of our analysis. If an event occurs or circumstances change that would more likely than not reduce the fair value of a reporting unit below its carrying value, goodwill will be evaluated for impairment between annual tests. No triggering events indicating goodwill impairment occurred during the three and nine months ended March 31, 2016.