Please note that there actual and assumed numbers.  Assumed or projected numbers are indicated by a (*) beside it.

I will try to go into some detail and explain how the projected numbers came about.  I didn’t do that in the previous thread, unlike my Sony threads.  Then, I will compare to see how “accurate” the prediction was.

This was the previous thread:

http://www.vgchartz.com/forum/thread.php?id=31432&start=0

Keep in mind that there is an inherent degree of difficulty predicting MSFTs EDD numbers, due to the fact that it was lumped with all other products.

Also, note that I actually did the MSFT projections before I did my Sony analysis (in my previous threads).  Therefore the MSFT projections and analysis were actually done with less care.  FYI.

However, in conclusion, I think you will see how the original prediction was quite close to the actual numbers that MSFT reported.

If any of you are a numbers geek, then let me know and ask nicely.  I may even send you the Excel file.  J  Cheers and happy reading and analyzing.

The original prediction was:

Revenue: 1,780(*) million

Profit: 100(*) million

Again, here was the original thread:

http://www.vgchartz.com/forum/thread.php?id=31432&start=0

Let’s take a look at Revenue:

This is where the BIGGEST mistake was.  (Note:  When I did the MSFT projections, maybe I wasn’t thinking too much or too optimistic, either or, BIG MISTAKE)

In projecting $1,780(*) in revenue, I projected 2.1(*) million consoles to be sold.  The actual number reported was 1.3 million consoles sold.  That’s a huge difference and basically explains the revenue variance of +13%

Using the EXACT SAME NUMBERS and assumptions, if I changed the consoles sold to 1.3 million, the projected revenue would be $1,501(*).

This would only represent an actual variance of $74(*) million or -4.6%, well within the 5% variance range

Let’s take a look at Operating Income:

The original operating income projected was $100(*) million – based on the same 2.1(*) million consoles sold.

In my model, if I use the 1.3 million consoles sold, the profit would have been $84(*) million.  Not bad, but still quite a bit different from the actual $188 million LOSS reported.

However, based on MSFTs reports, here are some other facts:

For the Quarter:

“Research and development expenses increased $141 million or 38%, primarily reflecting increased headcount-related expenses, increased product development costs, and costs relating to Danger, including a $24 million in-process research and development expense.”

For the Fiscal Year:

“Research and development expenses increased $242 million or 18%, primarily reflecting increased headcount-related expenses and costs relating to Danger, including a $24 million in-process research and development expense.”

How is this relevant in the analysis and projection?

Basically, in the most recently quarter, R&D increased $141 million, representing a HUGE increase in the Q4 alone, since the total for the whole year was “ONLY” $242 million.

If you go by just regular projections and use the same amounts as the previous 3 quarters, the “increase” in R&D in Q4, should have been a more modest $34(*) million.  This is what I took into account in my projections.

Therefore, the difference for the purposes of projection is $107(*) million ($141 less $34(*)).

Was this a one time increase?  Or is it a permanent increase in R&D expenditure?  It would be hard to tell.  We’ll see in the next quarters.

So now, the profit projection if you take this unexpected increase is a LOSS of $23(*) million.

Continued next post for easy reading...

Sales & Marketing Expenses:

For the Quarter:

“Sales and marketing expense increased $63 million or 22%, reflecting an increase in marketing and advertising campaigns and an increase in headcount-related expenses associated with product marketing and the retail account sales force.”

For the Fiscal Year:

“Sales and marketing expenses increased $93 million or 8%, primarily reflecting increased headcount-related expenses and increased bad debt expense.”

Again, I would argue this is somewhat unexpected.  The increase in the Q4 of the year represents 200% over the increase for the past Q3.  200% increases are generally quite unpredictable.

Using the same calculations as R&D, the allocated additional “expense” for marketing should have been about $10(*) million only.

This is a difference of $53(*) million ($63 million less $10(*) million) from the projected profit.

So now, the profit projection is a LOSS $76(*) million.

Danger Acquisition

The Danger acquisition most likely contributed a loss for the quarter as well.  In the FY ending Sep 2007, Danger made a revenue of $56 million and a LOSS of $28 million.

Even with an increase in revenue in the most recent quarters, most likely the company is still making a loss, as evidenced by R&D expenditure of $24 million attributed by MSFT. 

Not only that, the actual business operations of Danger is most likely losing money as well, as these types of companies (start ups) that are acquired by bigger companies usually are.

In this I project a small $15(*) operating loss for the quarter.

FIRST SUMMARY:

For MSFT Q4 FY ending June 30, 2008:

Projected Revenue with 2.1 million consoles = $1,780(*)

Projected Profit with 2.1 million consoles = $100(*) million

Projected Revenue with 1.3 million consoles (not including Danger) = $1,501(*)

ACTUAL Revenue = $1,575

Variance = 4.6% = PRETTY CLOSE

Projected Profit (Loss) with 1.3 million consoles = $84(*) million

LESS:

Unexpected additional R&D expenses = $107(*) million

Unexpected additional marketing expenses = $53(*) million

Additional projected Danger operating losses = $15(*) million

TOTAL projected LOSS = $91 (*) million

ACTUAL Loss = $188 million

Variance: $97(*) million = PRETTY FAR

So where did these additional losses come from?

More Additional Losses:

Based on MY model and assumption, it was actually quite easy where this came from.  These losses actually came Windows Mobile and from the X360 hardware.

Why?

For 3 quarters, Windows Mobile was allocated/projected to have had $100(*) million in losses.

However for Q4, I projected a $10(*) million profit.  I should have used a $33(*) million loss.  That is a difference of $43(*) million.

Don’t ask me why.  Like I mentioned in the beginning of the post, this was done with less care and hasty manner.

Also, for the past 3 quarters, my model allocated a net LOSS of $75(*) per X360 console.  However for Q4, I projected a $20(*) PROFIT per console – a huge and significant difference.  Maybe I was optimistic and reading too many MSFT fanboys claims, who knows?

But if I assumed a more reasonable $20(*) LOSS per console, instead of a $95(*) swing in hardware profitability, then that represented a change from $26(*) million profit to a LOSS of $26(*) million.  That is a difference of $52(*) million.

Again, don’t ask me why.  It seemed reasonable at that time.

SECOND SUMMARY:

Therefore:

Projected Profit (Loss) with 1.3 million consoles = $84(*) million

LESS:

Unexpected additional R&D expenses = $107(*) million

Unexpected additional marketing expenses = $53(*) million

Additional projected Danger operating losses = $15(*) million

LESS:

Difference in Windows Mobile projections = $43(*) million

Difference in X360 console projections = $52(*) million

TOTAL projected LOSS = $186 (*) million

ACTUAL Loss = $188 million

Variance: $2(*) million = PRETTY CLOSE

I hope this explains some of the thought processes in doing the analysis.  As we move forward, hopefully the analysis keeps improving and gets better over time.

Enjoy and Discuss.  Thank you for reading and give me some love on my thread.  Haha.

.

Was there a discussion on MSFT's results? I didn't see it in the main threads in front.