J.Edward

Accountant here.

The true raison d'être of GAAP is to allow meaningful comparison between companies.

Public companies in the US are obligated by law to report their financial performance based on generally accepted accounting principles, or GAAP. For companies outside the US a different standard of GAAP (as set per the country - IFRS might one day change this, but that day is not today) will apply.

GAAP works through standardization of information; specifically the standard treatment of business events (sales, expenses, losses, etc.) and the reporting thereof, and while this often works well to facilitate comparisons between companies issues can arise. Specifically the more complex an industry is - and aviation is no exception - the more GAAP struggles to standardize the results.

Picture in your mind a clothing store that only sells one size of suite and then picture that all executives were required to wear it without alteration or adjustment. You would see a harmonization in that everyone is wearing the same sized outfit but the fit would range between good to atrocious/comical as the executives would differ in size and shape based upon the nature of who they are.

This is both the strength and weakness of GAAP: it facilitates a framework for standardization but standardization is not a end in itself, it is a means to an end! Remember GAAP's purpose: to help investors understand the health of a business through comparing it to others business. It does not exist to standardize information for standardizations sake nor to file reports for reports sake.

Hence the "G" in GAAP: GENERALLY (as in NOT always)

In our one-sized-suit example if company performance = how well the exec looks, then there is some comparability that can be ascertained. Let's say all the airline execs were very tall with extremely short arms, then the same sized suite would fit all of them poorly and an investor at first glance might assume they are all performing poorly as they look atrocious in their GAAP suite.

Granted you could look to see which suit fit the best (or least worst) to compare the exces but ultimently bits of who they are would not fit well into the one-sized-fits-all approach as the one-sized-suit is making certain aspects look worse than what they really are, and that goes against the main reason of GAAP: to allow investors to accurately see the company (or how well the executive looks).

Hence you have the non-GAAP argument - this is akin to the standardized suit being tailored to shorten the arms but lengthen the legs. Granted a deviation has been made, but since better reflects the nature of the industry (as all the airline execs have short arms and long legs and they all do the alteration to an extent), you can see a better picture of the exec and compare them with their piers without a (at times) poorly fitted GAAP suit getting in the way.

As each industry will have their own quirks which are contradictory to GAAP an argument can be made to electively deviate from the GAAP treatment of items in favor of a non-GAAP approach, assuming there is a valid reason (e.g. doing so better reflects the true nature of the transaction, and thus presents a more accurate portraly of the company to the investor).

Currently if a company deviates from GAAP they have to list why (see below for UA's reasoning for doing so), present the GAAP numbers along side the adjusted numbers and provide a reconciliation for the non-GAAP to GAAP numbers.
This is incorrect.

While GAAP literature can be rather vague on defining special items such as non-reocurring vs. extraordinary (and IFRS does not even recognize an extraordinary item!) a key theme is that the special item is not reoccurring.

There are other nuances in how these will play out but in the case of UA accelerating the phase out of 145s (or DL with 747s), a clear argument can be made that this will not be a reoccurring expense.

It should also be noted that UA's continued catorigaztion of "special merger changes" into this year raised eye-brows (IIRC Jamie Baker (?) asked point blank when the charges would be over on the Q1 CC) as people questioned if the charge should be considered "special" considering the ongoing history of it.

Finally you can also have special items that are anticipated - for example UA might be planning to accelerate the retirement of their 747s in Q1 2015 but for a variety of justifiable reasons might not wish to share this information until they commit to it. Personal changes/layoffs/buyouts could also fall under special items (assuming they are considered non-reocurring) but they certainly may be anticipated.

Anyways the reason I spelled all this out is to address a few straw-men that seemed to be raised earlier in the thread:
Glad to see UA finally having a good quarter - hopefully they can keep up the momentum! ^