Wednesday, March 21, 2012

Silverlake Axis - Part 3 (SIBS, The Source of All Profit)


Fig 1 Revenue by Segment

In this part 3, we will focus on how Silverlake Axis generates its revenue through its propriety software, Silverlake Integrated Banking Solution (SIBS). As such, I will only be touching on "Licensing of SIBS" and "Maintenance and Enhancement Services", which are its core business. The rest of the other businesses will be reserved for part 4.
For licensing of SIBS, it is a fairly simple business to understand. When a bank has decided to change its core banking system, the first step will be to build or buy? In the 1990s, many have tried to build a system themselves and this has proven to be a terrible and costly mistake. As such, in the current era, many especially the smaller banks will seek out a vendor to provide the solution. This is a very tedious process that will usually take a few years to complete and will be even more complicated should the bank has many branches.
This is a highly lucrative project whose profit margin has always been at least 90%. SIBS has already been developed as such the bulk of the cost comes from the software engineers that are hired to install and implement the solution. How much the total contract size will be is determined by the number of branches that a bank has and the type of business it is being involved. Thus, so long as its existing customers continue to expand the number of branches, it will be able to earn additional software licensing fee. However, in normal circumstances, the bulk of the revenue comes from new customers. Therefore, we need to recognise the fact that this source of revenue is contract-based, lumpy and usually one-off. This is also the main reason why Silverlake Axis has consistently been poorly valued, be it too high or too low. But the end of an implementation contract does not signify the end of all working relationship and profits, instead it just marks the start of an additional source of recurring income for as long as the bank continue using SIBS. (remember in part 2, it was concluded that the vendor and the bank are likely to stay engaged for at least the next 15 years?)
After the successful implementation of SIBS, a 6 months- 1 year warranty period is provided for the bank for free. After the end of the warranty period, a "maintenance" contract will be signed where the vendor will provide maintenance service to the bank for a fee. While it is called a "maintenance" contract, in reality, it is simply a form of license fee to continue using the software for the bank. So how are fees being charged? The industry standard for this annual maintenance fee is 15-22% of the original licensing fee for the software, which is what that makes the industry so lucrative. Or to put it in another way, the vendor will get to earn another full licensing fee after another 5-6 years. And of course, if the bank expands and incur more licensing fee, expect the maintenance fee to increase too. This is just the tip of the iceberg. SIBS will have a new global upgrade version after every 3-5 years and the last version was out in 2010, which means if a bank wants to upgrade the CBS version, it will have to pay another amount. (One can choose not to upgrade though)
http://info.sgx.com/webcoranncatth.nsf/VwAttachments/Att_DF2B24500EF455F8482577BB00205A3F/$file/Silverlake_Announcement_SIBS_Upgrade_Contract.pdf?openelement
I supposed this large regional bank should be none other than our OCBC, there's not many regional banks around whose upgrade contract can cost $30m...
With such lucrative business, will Silverlake Axis be capable of fending off its competitors, especially with the like of Tenemos, Misys, Infosys and Fidelity? Will it be able to fight against them to secure new contracts? Before I touch on this point, lets look at its existing customer base.
Singapore - OCBC and UOB. UOB uses their software since 1997 and has in fact been one of their most important client in their history. UOB was using an advance CBS from a US vendor but has decided to replace it with SIBS. This success allows it to secure further contracts from OCBC, May Bank, Hong Leong and Mandiri. To conclude, 2 out of 3 local banks in Singapore uses them. In case you are wondering what software DBS has been using that has been creating the most number of glitches in Singapore, they have been using Infosys Finacle since 2005. They even have a write-off during the implementation
http://www.ibsintelligence.com/index.php?option=com_content&view=article&id=12919:dbs-project-in-doubt&catid=227:2009&Itemid=103

Figure 2 - 8 Anchor banks in Malaysia

Malaysia - This is in fact its strongest ground given that it is a Malaysia company. 6 out of 8 anchor banks in Malaysia uses SIBS (those highlighted in Orange). Bank Islam the number 9 largest bank in Malaysia also uses SIBS, they are the oldest Islamic Bank in Malaysia. (Please correct me if I am wrong) Other customers include Employees Provident Fund (our equivalent of CPF) and Malaysian Airline System (their national carrier).
Fig 3 - Top 10 Banks in Indonesia

Indonesia is yet another stronghold of Silverlake Axis. The banking industry in Indonesia is very fragmented with more than 200 banks due to the low entry requirement to open a bank. Among the top 10 banks, 4 uses SIBS - Mandiri, Bank Rakyat Indonesia, CIMB Niaga and BTN. Coincidentally, Mandiri, BRI and BTN are 3 of the 4 state-owned banks (the last one is BNI). Just these 4 customers account for 32.1% of market share in Indonesia. Silverlake's other customers include - Mega (11th), OCBC NISP (13th) and Buana (14th).
Vietnam - Vietcombank (3rd largest), BIDV(2nd largest), IncomBank (4th largest now known as vietinbank) and Maritime Bank (6th largest). Vietcombank, BIDV and Incombank are 3 of the 5 state-owned banks in Vietnam.
Brunei - Bank Islam Brunei Darussalam Berhad (BIBD), Perbadanan Tabung Amanah Islam Brunei (TAIB). TAIB and BIBD are 2 of the only 3 local banks in Brunei
Sri Lanka - People's Bank, largest in term of network and 2nd largest in Sri Lanka. It is once again another state-owned bank.
Philippine - UCPB (10th largest), Bank of Commerce (13th), PBCom (19th), Robinsons Bank (23rd, it is the largest thrift bank too) and Premiere Bank
Thailand - Thanachart Bank merge with Siam City Bank to create the 5th biggest bank in Thailand.
HSBC Amanah - Largest Islamic financial Service from an International Bank. 8th largest non-government Islamic Finance Institution.


Fig 4 HNA Group
China - Commercial Bank (9th largest), HNA Group. Global Infotech (22% stake in it) - 90 clients in China, notably 2 state-owned bank - China Agricultural bank and ICBC. But what's the most interesting here is its partnership with HNA group.
As seen above, HNA group is a big conglomerate that involved in retail, finance, airline, property, shipping, airport, logistic and e.t.c. It is a rather mysterious group due to its size and scope, but is not a state-owned enterprise. It is one of the 7 airlines to be ranked as a 5-star airline globally and have total asset of 160 billion RMB. Controlling numerous airline, it forms the Grand China in 2007, with George Soros having a 18.6% stake in it. HNA has numerous substantial stake in Beijing Capital Airlines, Lucky Air, Zest Airways, Tianjin Airline, Yangtze River Express, Hong Kong Airlines and Hong Kong Express Airways. Why did HNA Group chooses Silverlake Axis- because they are able to provide software solution for finance, retail and airline as well as in card processing.
The SGD 70m contract announced in 2010 is fortunately unlikely to be the last contract that they will secure from HNA Group. http://info.sgx.com/webcoranncatth.nsf/VwAttachments/Att_168CB7207F663A2F4825775B0030DABB/$file/SAL.HNA.Strategic.Alliance.Annoucement.9July2010.pdf?openelement
Shortly, in June 2010, Mr Goh Peng Ooi entered a structured share sale agreement with HNA Group where a total of 11.57% of issued shares will be sold by Mr Goh to HNA Group over a 3 years period at a price of $0.16. The price of $0.16 will be fixed should the price increases, but if the price drops, Mr Goh will then have to sell those shares at the lower price instead. For Mr Goh to undergo what seemed like a short-changed deal, obviously this strategic partnership is going to develop to much more contracts in the future or perhaps something more than just 11.57% shareholding.
From the above, it naturally gives us one conclusion - Silverlake Axis has a very strong position in the South-East Asia region especially Singapore, Indonesia, Malaysia, Brunei and Vietnam. Let's have a look at some statistical evidence.


Fig 5 - Inntron Core Banking System Ranking

Fig 5 is a ranking of the top 10 CBS vendor globally. From the above, we can see that there's 5 USA firms, 2 India firms, 1 UK firm and 1 from Switzerland. For Silverlake Axis to get into the top 10 in the world is very impressive for a firm that has usually been operating only in the South-East Asia Region. Of course, there is no point being in the top 10 if 9 other competitors are ahead of you.


Fig 6 Ranking of Islamic Banking Solution
Silverlake Axis comes in at an impressive number 3 position, which is not surprising given its strength in Malaysia and Indonesia.
Fig 7 Top 10 in Asian Regional Market
Silverlake Axis is ranked number 4th in the whole of Asian regional market. From the above 3 rankings, it seemed like Silverlake Axis is not that strong in Asian and Islamic Banking, with Temenos and Oracle consistently coming in front of it.
What I believe is Silverlake Axis's true competitive advantage lies in South-East Asia, Islamic Banking as well as in small-tier and mid-tier banks. For South-East Asia, it is in fact very clear that there's no other vendor whose operation comes close to it. The closest competitor in this area is Infosys's Fidelity. Being stationed in Malaysia and its 20 years track record of successful implementation in major banks in Singapore, Malaysia, Indonesia, Brunei and Vietnam, these explain its stronghold in South-East Asia. Other than this, it is able to integrate its Islamic Banking solution with its SIBS, as well as providing other services like card processing for the like of Bank Rakyat Indonesia.
Another point to note is that Silverlake Axis will never implement a full scale solution for the likes of HSBC, Barclay Group, CitiBank and Deutsche Bank. Silverlake Axis has always been focused on small bank and mid-tier regional bank like OCBC, UOB, Maybank and CIMB.
And the question now will then be why should one be invested in Silverlake Axis and not Oracle, Fiserv or Temenos? The answer again lies in its stronghold in ASEAN. Among the global banking industry, banks in South-East Asia has one of the strongest balance sheet to withstand any financial crisis after having been through the 1997 AFC. They are also poised for growth with the banking scene of many like Indonesia and Vietnam being undeveloped. For the more developed one, regional bank like Maybank and CIMB are actively acquiring and expanding. As noted above, an expansion of branch will result in additional licensing fee and hence the Maintenance and Enhancement fee as well.
Countries like Malaysia and Indonesia are also liberalizing their banking sector and hence consolidation is expected. In any M&A, it is very likely that the acquirer will have its CBS imposed on the acquired. This can be a double-edged sword for Silverlake Axis, because if its customers got acquired, it may lose its maintenance fee. However, with the biggest banks in the region as its customers, it is more likely to gain more customer than to lose them. Last year, Silverlake Axis was so close to getting its hand on the 5th biggest bank in Malaysia, RHB, in a deal that reminds many of the consolidation happening in Singapore in 2001. Banks in Malaysia and Indonesia has been very aggressive of late.
http://www.asiamoney.com/Article/2848330/Malaysias-top-banks-fight-in-final-bout-of-M-A.html
There's 3 points that we have established in this post
  1. CBS is a highly profitable deal, with a 90% profit margin for the initial licensing fee, followed by a generous sum of recurring income for as long as banks continue using the software.
  2. Silverlake Axis is a market leader in South-East Asia, no other vendor is anywhere near it.'
  3. Hence, it will get to leech on the growth of the South-East Asia banking industry as well as the liberalization going on in Malaysia and Indonesia.
Part 4 will be on All Other Businesses of Silverlake Axis like SBI Card Processing, sales of hardware, Unisoft, QR tech and e.t.c

9 comments :

  1. Hi, interesting read. I have a few comments though.

    1) Are you certain that banks that use SIBS do not use other banking platforms as well? I seem to recall seeing data files from both SIBS and Midas (by Misys) whilst working in one of the aforementioned Singapore banks. I got the impression that different divisions in the banks may choose to use different platforms based on their preference. If this is true, it seems to weaken your Silverlake stronghold theory with respect to SEA banks. They would be just one of the vendors instead of the only vendor. Every existing vendor would have a working relationship with the bank, thus no one has a real "leg up" on the others.

    2) Pertaining to HNA Group, my interpretation is slightly different from yours. Firstly, looking at the 2011 AR seems to indicate that HNA Group owes a large trade receivable to Silverlake of about RM44mil and the shares "sold" are transferred to the escrow account when each milestone progress payments fall due and PAID BY HNA.

    I interpret this as Silverlake offering a "free option" to HNA in exchange for their good behaviour, i.e. pay up on time. According to the text in the SGX announcement, "At exit through SALE IN THE MARKET or BACK TO IHL, HNA will enjoy all gains above S$0.16, with Mr Goh Peng Ooi receiving S$0.16". The fact the Mr Goh gets back S$0.16 upon any subsequent sale shows that there is no ownership transfer when the shares are deposited into escrow. HNA has no skin in the game and loses nothing if Silverlake goes down in flame. However, HNA gains if Silverlake's share price perform well. This is a rather assymetric risk/reward. On a positive note for Silverlake, this agreement will encourage HNA to give Silverlake more business in China.

    3) The large gains in the lumpy non-recurring categories of "Licensing of SIBS" and "Customised Software Solutions" are the main reason for the large jump in revenue and profit in 2011. "Maintenance and Enhancement Services" is stable and slow growing historically, and should continue as such going forward. As you mentioned in an earlier comment, Silverlake is more like a 20x PE company (if we normalised income). If my comment in point 1) is true, would you reconsider whether Silverlake is a good buy at such a high price multiple? After all, recall that in 2009, the original (not restated) income statement showed a large decrease in Silverlake revenue and earnings when banks tighten their belts during the GFC. Banks are cyclical creatures by nature. They experience great earnings growth during economic recoveries and large credit losses when recessions hit. Emerging market banks are double-edged swords. They tend to grow fast, but may have lax credit risk management. With all this in mind, would it be better to wait for a better margin of safety, i.e. lower entry price?

    ReplyDelete
    Replies
    1. 1) This could be likely because that the bank does not wish to change the CBS of those that it acquired. It is not neccessary, but it will just make the bank more inefficient. May I know which bank are you talking about? are you talking about OCBC and Misys' Treasury & Capital Markets solution called "OPICS". That's a system for money market, forex and e.t.c

      And as far as I know, Silverlake's core competency lies in retail banking solution and all its client are mainly using it for retail banking purpose. This is why it is more catered to the small-tier and mid-tier bank. For bigger international bank, they often have multiple platform to cater to different division - Wealth management, capital market, retail banking, investment banking. However, the essence of CBS will still be in retail banking as that's where account transactions of the masses are.

      2)Indeed, HNA group owes Silverlake a large amount of receivables. However, the agreement is not for the receivables. If you notice, the agreement was made 5 months after the contract was announced. Within this 5 months, I doubt much implementation was being done and therefore it sohuld not be about receivables. I tend to view this more positively as no dilution was being done and it was Mr Goh himself that suffered from this deal. Therefore, I do believe that this was more for the sake of partnership.

      3)There is never a limit for Margin of Safety - it is always the more the merrier. I did ask myself then, what was my margin of saefty when i bought this company given that I am buying at a P/E of 20. I will give my reason for buying at this P/E in my last post. For now, just remember that at least 15% of the original licensing fee will go into mainenance fee yearly 6-12 months after the successful implementation.

      As for it being cyclical, you should check out the 2010 annual report, where the 2009 was being done pro forma had the acquisition happened in 2009. Then, you will have understand whether the industry was cyclical. Compare it too with the original 2009 annual report and you will have understand how important the acquisition was.

      And i have just been informed by a reader that Silverlake has just increased the maintenance fee by 10-15% across the board according to a report by DBS.

      You seemed to be working in the bank, I really appreciate your input. Hope you can share more about it

      Delete
    2. Yup, was referring to OCBC. Was not in IT department though so can't really add much besides end-user laymen views.

      I see ur point abt licensing leading to greater increase in "Maintenance and Enhancement Services" revenue. 15% of "Licensing of SIBS" RM90+mil revenue in 2011 brings additional RM14mil maintenance revenue, more than 10% growth. If Silverlake is able to increase the maintenance fees, all the better.

      I was referring specifically to the banking industry about being cyclical, not Silverlake itself. I guess the acquisition in 2009 does make it a stronger company due to the maintenance contracts. However, new business probably remains hard to come by in times of banking uncertainties, and may somewhat impact earnings. That said, no company is perfect, right? 

      Delete
  2. I think there's some confusion here. The IT model of most banks is based mostly on one core banking system other than international banks. Misys's opics is just a front-end software and not a core banking system. It is more like an add-on and any other software like Adobe Reader, SAP and so on. You can have many software, but it will likely be that there will only be 1 core banking system

    ReplyDelete
    Replies
    1. another banker here, you are welcome to share your input. You are right, I will treat CBS as something like Window OS while you are free to install any other software that can help you in your task on your computer

      Delete
  3. Shall look forward to your final part. Would love to hear how you substantiate an entry PE of 20

    ReplyDelete
    Replies
    1. No worries, I have not ventured into the dark side of speculation, at least not for now.

      Delete
  4. Hi 20yo Investor,

    I would like to ask, what is the view on the HNA structured share sale agreement? HNA buys a portion of SAL (owned by IHL), and how does SAL directly benefit from the sale?

    Also, do u know what are the milestone agreed?

    ReplyDelete
    Replies
    1. The HNA structured agreement does no harm to Silverlake as it is Mr Goh that sells his share. Hopefully, HNA will be more inclined to provide Silverlake with more contract in China.

      http://info.sgx.com/webcoranncatth.nsf/VwAttachments/Att_168CB7207F663A2F4825775B0030DABB/$file/SAL.HNA.Strategic.Alliance.Annoucement.9July2010.pdf?openelement

      i believed it should be split equally among three years

      Delete