Reflection on Silverlake Axis

Sooner or later, the markets always humble people. I believe in it and have my fair share of being humbled. While I was not hurt in monetary terms as I sold it off 2 years ago on price discipline, it will be good to do a personal reflection.

http://sgyounginvestor.blogspot.sg/search/label/Equity%20Research%20Report%20-%20Silverlake%20Axis

My mistakes were

1) I failed to pay attention to the related party transactions
2) The past deals action did not come across as red flags to me as I thought I had invested in the new entity where the whole Silverlake Group is inside the listco.

I tend to be bias to the short seller and will be glad to have more short report in the market as it improves the market process and mechanism. Who is razor99? razor99 called a short on Longtop months before Citron launched a successful attack on the company. Longtop specialises in software catered to the financial institutions, an area similar to Silverlake. Both his experience with Longtop and the release of a fraudulent claim in May probably led razor99 to look into Silverlake.

The most damaging claim

Razor99 made many claims in his 42 pages report, including concerns of bribery. However, only 1 claim was truly critical. According to razor99's claim, margin at Silverlake was inflated through related party transactions. The evidence was a jump in revenue and margin at to-be-acquired entities. Another evidence was the losses suffered at Goh's private co such as Sprint.

Legally, it might be hard to accuse Silverlake of fraud as the company has followed the proper procedures:
1) The mandate for IPT was approved by shareholders annually since 2008 and had always been disclosed since the 2003 IPO
2) The details of the IPT was disclosed in the annual circular (inclusive of pricing)
( http://www.silverlakeaxis.com/investor-relations/financial-information/circulars.html )
3) Auditors will review every IPT transactions every year

What are the related party transactions (RPT)?

The IPT mandate approved by shareholder annually holds the critical information to understand the RPTs. In addition, lets take a look at Silverlake listco after 2011 since that is the current corporate structure.

Source: razor99 report

The bulk of the RPT revenue is from software licensing while the bulk of the RPT expense is from service fees paid.

RPT Revenue % per segment

To put the information in another perspective, it will be to view the segment revenue and determine how % of segmental RPT revenue against the actual segment revenue. It can be seen that the bulk of the software licensing revenue (more than half of it) comes from RPT. This deserves a deeper look since software licensing has the highest margin (90%) and account for at least 40% of Silverlake's profit.

Master License Agreement

Everything is properly accounted for under the MLA, inclusive of the minimum fee of US$20 million and how the points are allocated. Under the Master License Agreement (MLA), the interested person is granted the right to "resell, implement, copy, customise or use the software" and to "sub-license the right to use the Software to End-Users". Essentially, when the interested person wins a new software contract, Silverlake will earn a license fee while the implementation and employee costs are borne by the interested person. This explains the 90% profit margin for the software licensing revenue. You don't expect to pay any cost other than your R&D amortization and expenses.

Master Services Agreement

Under the Master Services Agreement (MSA), this deals with the receipt of "Customisation, Implementation and Maintenance services by the Group from the Interested Persons". The MSA explains the 50 million service fees paid, which is almost all of the expenses paid out to the related party. With the MSA, Silverlake sub-contracts or outsources part or whole of the implementation and maintenance job to the interested party for a fee. Once again, everything is properly accounted for with the use of "Man-Day Rates" and "estimated man-days required to complete the work"

Peer Analysis, Source: razor99's report

With an understanding of the MLA and MSA, it is a lot clearer why the peer analysis results in Silverlake having a low employee expenses %, much higher profit margin and much higher revenue per employee. A huge portion of a new software contract implementation is outsourced to the Interested Person. An unknown portion of the costs for maintenance and enhancement is also outsourced to the Interested Person.

Legally correct but still a governance concern

Being right legally is not equivalent to being right in principle to the shareholders though it is highly debatable in this case. Inevitably, the question in shareholder's mind will be what kind of profit is Mr Goh's private entity making from these MLA and MSA contracts. 

1. If it is a loss, Mr Goh runs the risk of being accused of hiding expenses and inflating margin (although everything is properly and legally accounted for, so it is a false accusation).
   
   And a subsequent question will be how long is he willing and able to sustain the loss on his own account? Ans: For as long as he continues to be generous. (With regards to the ability, high dividend payout creates a feedback loop and share sales can top up the remaining amount)
2. If it is a profit, why not share it with the shareholder to show an alignment of interest?
3. Even if it break-even, why not consolidate all the private co into the listco and nobody can accuse you of anything in the future?

There are of course numerous clues that point to 1 particular scenario above if one read razor99's report and understand what have been written so far on the MLA and MSA.

In any case, for proper corporate governance, it is still best to consolidate the private co into the list co as the Nash Equilibrium. Shareholders can recommend to Mr Goh and they also have the right to vote against the IPT.