P/E Ratio - Which One Is Correct?, page-7

I answered this in a reply to your question when you asked it two days ago. My answer was:

As discussed several times previously here, that Commsec PE is not really accurate, it's based off the statutory FY17 NPAT which included a one-off amortisation hangover from the Dick Smith acquisition. [edit: the main cost is actually the IPO costs, not the Dick Smith amortisation]

Using the pro forma NPAT of $7.2M, trailing PE is about 85. Using the 1H18 NPAT of $8.3M and annualising it (ie if we assume no NPAT growth for 2H vs 1H) the PE is about 37.

If NPAT grows ~30% in 2H vs 1H, then PE would be about 30. Still expensive, but nowhere near 180, and not ridiculously expensive for a stock with that sort of NPAT growth.

So realistically the PE is probably 30-40 right now.

SP has dropped further since I posted that so PE is about 29 based on half year results.

The big difference between this number and the PEs listed on Bloomberg and Commsec is that they use the last official full year earnings which is the year ending June 2017. They won't update their EPS until the next earnings are released next month, for the FY ending June 2018.

The difference between Commsec and Bloomberg seems to be that Commsec is basing its calculation on a "NPAT before abnormals" number of $3.3M to get an EPS figure of 3.6 cps (as you can see in Financials section of Commsec's KGN page), while Bloomberg is basing its calculation on the statutory NPAT of $3.7M to get an EPS of 4.0 cps. I'm not sure how Commsec got that $3.3M figure. Bloomberg is more accurate.

Regardless, the statutory EPS for last financial year was 4.0 cps ($3.74M NPAT divided by 93.3M shares). But as I said, this includes a lot of one-off costs so the pro-forma EPS is almost double that, and the EPS for the 1H18 was 8.9c just for the half year.

It's a good example of why you should never just look at the PE number in isolation, especially for a stock with volatile earnings like KGN.