Monday, July 22, 2013

SPH REIT - A Lesson on Financial Engineering 101

SPH REIT is a spin-off from SPH, comprising of Paragon Mall and Clementi Mall, offering an annualised yield of 5.58% for FY 2013. At this income supported yield, it offers a 0.5% higher yield than CMT and 0.2% lower than Fraser Centrepoint Trust. Retail REIT has always been a favourite of REIT investor due to their defensiveness and ability to squeeze extra rental cost from the tenant.

However, why is it that they delayed their listing on 24th June 2013 before lodging their prospectus on 9th July 2013? In a mere 3 weeks, an unattractive IPO is now the hottest thing on the market, with an indication of interest from institutional investor being 42x the placement tranche. http://www.stproperty.sg/articles-property/singapore-property-news/sph-reit-attracts-strong-institutional-interest/a/128627 . At the end of the day, many investors define the attractiveness of a REIT based on the yield. This prompts me to look at what forms of financial engineering have been used to increase the yield. Shall start the analysis with the most obvious and end with the hidden trap. The hidden trap is what that allows SPH REIT to be priced at 0.5% higher yield than CMT.

99 Years Lease. Similar to K-REIT, SPH sold a 99 years lease of Paragon Mall to SPH REIT while retaining the freehold ownership of Paragon Mall in our prized Orchard Road. According to the Knight Frank valuation certificate on the circular by SPH, Paragon mall is worth $2.61 billion as a freehold property and $2.5 billion as a 99 years leasehold property. The Freehold Lease is worth only $110 million, which is ridiculous. Who will not want to pay an extra 4% of their property value to upgrade their property from a 99 years lease to freehold? 99 years lease represent a 1% depreciation each year, which means the payback is theoretically 4 years to secure the freehold right. Of course, it is likely that maintenance and capex will be needed for an asset past 99 years.

Income Support. SPH has guaranteed that Clementi Mall will produce $31 million in NPI for the next 5 years. The reason is that Clementi Mall is a relatively new mall and that the current rent signed is lower than the market rental. Income support is common for such new property, but does SPH really lose money from providing income support? According to the valuation report in the prospectus, the $570 million Clementi Mall is only worth $550 million without the income support. Projected NPI from Clementi Mall is $26.7 million in FY 2014. On the conservative assumption that it remains the same for the next 5 years, SPH will provide a total income support of $4.3 million x 5 = $21.5 million. Thus, SPH is unlikely to have to fork out more money by providing the income support on a net basis (it earns extra $20 million from selling at $570 million). Yet, investor can be sold on a higher dividend yield though it is a partial capital return of the IPO investor. In addition, given the income support for 5 years, it means that growth for Clementi Mall will already be factored in for the unitholders unless they can grow their NPI beyond the income support mark of $31 million.

Lower capitalisation rate in valuation. The gearing ratio of 27% looks attractive and helps to alleviate fear of rising interest rate. However, gearing ratio can be engineered through a change of the property valuation which is the denominator. A cap rate of 4.75% and 5% have been used to value Paragon Mall and Clementi Mall (the income support has been taken into account) respectively. In comparison, CMT uses an average cap rate of 5.4% to 5.85% while FCT uses an average cap rate of 5.50% to 5.75% to value their properties. Some might argue that as 99 years leasehold asset, they should be worth a much higher valuation than the 70+ or 80+ years remaining leasehold asset. However, Atrium@Orchard with 94 years remaining lease and Yee Tee Point with 92 years remaining lease do not have a lower cap rate. This reflects how conservative or aggressive a management has been in valuing the property and thus tweaking the gearing ratio.

Source:

Lower interest rate. According to the prospectus, SPH REIT has been assumed to pay a constant 2.35% interest rate on its loan. The $850 million loan is to be repayable 1/3 each in 3,5 and 7 years time which gives an average loan maturity of 5 years. s of 31 March 2013, the effective interest rate for FCT and CMT are 2.73% and 3.3% on average term to maturity of 3.35 years and 4 years respectively. How did SPH REIT manage to secure a lower interest rate of 2.35% and higher term to maturity of 5 years when the fear is that interest rate is going to rise in the near term? Firstly, I cannot find the exact interest rate charged on the loan in the prospectus and the 2.35% has been quoted as an assumption. Thus the actual interest rate might be much higher than the assumed 2.35%.

Even if the assumption is true, how did SPH REIT do it? They actually got a secured loan over Paragon Mall which accounts for 80% of their total value. CMT's unencumbered assets as % of total asset is 76.7% and FCT is 43% compared to 20% for SPH REIT. (Unemcumbered means the asset has not been mortgaged. A secured loan is cheaper than an unsecured loan. However, when financial crisis comes, an unsecured asset might be what that will give the bank the confidence to lend you money. Thus, by taking a secured loan up to 80% of their total asset, SPH REIT is taking a higher refinance risk as compared to CMT and FCT just to inch out a lower interest rate and hence 0.19% higher dividend yield. On the other hand, it is likely that CMT will not face much refinancing issue given their unencumbered assets and well-spread debt profile.


The Finale

This is the most important point to take note of on how SPH REIT actually financial engineered a 0.5% higher dividend yield than CMT using an old, legal and similar technique of option expensing.


The Manager's management fee is $15.2 million in FY 2014 and they have chosen to be paid 100% in units. This is equivalent to taking a stock option, which convert employee's cash expense into 0 though it actually results in dilution in EPS the subsequent year. The $15.2 million to be paid in units is equivalent to 11.6% of the Income Available for Distribution. In contrast, the proportion of management fees paid in unit as percentage of Income Available for Distribution is only 1.5-2.5% for CMT and FCT.


Looking at the DPU sensitivity yield above, if SPH REIT were to take 100% of management fee in cash, the annualised yield for SPH will drop to 4.93% and 5.15% for FY 2013 and 2014, which will price it at a lower dividend yield than CMT. This is but a financial engineering move as the dilutive effect and selling of shares will weigh on the share price in the long run. The reason why the management fee is such a high percentage of distributable income is because SPH charges 0.25% of asset value and 5% of NPI as a REIT Manager and 2% of Revenue and 2% of NPI as a Property Manager.

To conclude, it is certainly not fair to simply compare the dividend yield of the REIT and determine how attractive it is. As always, the devil is in the details and I believe the market is a weighing machine in the long run.

24 comments :

  1. I like your analysis, impressive...

    ReplyDelete
  2. Agreed. Will need to follow you more closely on future IPOs to see how they should be analysed.

    ReplyDelete
    Replies
    1. thank you. i will if there is time and there's interesting ipo

      Delete
  3. Thanks a lot for putting in a 'financial engineering' perspective to this. I wonder if you have done a similar review for OUE R? It will be interesting to see if any engineering work was done to it.

    ReplyDelete
    Replies
    1. hospitality reit is more complicated than retail reit and is not within my circle of competence.

      i think the main problem should be the short land lease and that there's only 2 properties like SPH REIT. yet, hospitality reit derives more variable rental as % of the operator's revenue and operating income. During tough time, occupancy rate and revpar will fall which makes the distribution less stable and predictive. Thus, investor should demand a higher yield to compensate for the faster depreciation (due to short lease term) and the inherent fluctuation according to economic condition

      Delete
  4. Very good analysis. For laymen it is heartening to know that you have broken down the hard to understand details to more or less easily digestable facts. I hope you can also add some basic examples in explaining some of the financial jargon (even if it is in a separate article) so that many people not necessarily the elderly/retired share owners will be educated to understand future IPOs better...and hopefully better judge and invest. Thank you

    ReplyDelete
    Replies
    1. I will consider your suggestion. ideally, one should only invest if it is within his circle of competence. Investing in something that you don't understand is very dangerous

      Delete
  5. like your analysis. Could I ask since you are so good in fundamental analysis, could you shed some light on your portfolio return? Thanks.

    ReplyDelete
  6. Just to add on. what I mean is the returns that you wrote on your blog ending 2012. Is it based on XIRR or CAGR?

    ReplyDelete
    Replies
    1. thank you.

      it is based on CAGR. In any case, isolated investment return is insignificant. There will only be a meaningful purpose when compared against the benchmark index over 1 market cycle (for e.g. trough of a bear market to the next trough of another bear market). This will probable take at least another 8-10 more years.

      Delete
  7. Hi,

    Regarding the last point about financial engineering, could you elaborate a little more?

    The management fees were added back into the distribution income, and the total distribution units increase as result, how would this impact the yield?

    For example management fee is $10, and distribution income is $50 (before adding the $10 back). There are a total of 10 units that the $50 will be distributed to, so each unit receives $5.

    Now we add $10 back to $50 = $60, and subsequently, distribution unit also increases from 10 to 12 units. Each units still receives $5.

    Is this what that is happening? I'm not familiar with REIT, so any explanation / clarification would be helpful.

    Thanks!

    ReplyDelete
    Replies
    1. Guess you miss out one variable called the share price and no of share

      Let say that the share price is $5 and there are 100 shares in the market giving a market cap of $500. net distribution income is $50 and management fee is $10.

      If management fee is not claimed as units, dividend yield will be 10% (50/500).

      If management fee is claimed as units, the $10 management fee will give the asset manager 2 shares. Total dividend yield will now be [$60/(102*$5)] = 11.8%.

      Does this means that it is best for management fee to be claimed as units? Personally, I feel that this is a last resort move by management to increase the dividend yield when bad times hit. Thus, the main point is to compare against the peers to ensure that the dividend yield has not been falsely inflated against the peers. For e.g. it will not been fair to compare dividend yield of CMT and SPH REIT given the disparity in amount of management fee claimed in units. When the bad times hit (for e.g. interest rate goes up), CMT will be able to stabilise its dividend yield by using this last resort measure whereas unitholder of SPH REIT has to suffer a cut in dividend yield as the last resort has been utilised.

      Delete
    2. if your income has gone up from 50 to 60, and assuming the market is perfect (meaning price = value), your price should increase to reflect the higher income.

      unfortunately the market is usually not perfect, but that would be a separate discussion on price versus value.

      so the yields discussion is not entirely made on a comparable basis.

      Delete
    3. you are right that market will adjust accordingly to reflect the price at income of 60. But the price matters as it determines the number of shares that the management will get in lieu of management fee and hence the dilution.

      Delete
  8. as a young investor when you were 20, which bank account or service did you go to for your purchasing of sg listed stocks? and do they charge fixed price per trade or is it commission-based? thanks alot

    ReplyDelete
    Replies
    1. ocbc, cimb and philip accept ppl below age of 21. Commission charges is similar across banks with few exception like DBS upfront and standard chartered

      Delete
  9. Just a wild guess but looking at your portfolio & the way you analysis companies, must be an analyst or student from 8I, MIP. =]
    Like your work though. Lots to learn.

    ReplyDelete
    Replies
    1. Sorry, but I am not. If any, I am a student of valuebuddies. haha

      Delete
  10. I always have this lousy feeling that REITs are like scams by developers/sponsors to hive off their assets at inflated prices without much value add. They retain control of the assets and charge a new layer of management fees, which puts pressure on rental rates. Can you see significant value add by any of the existing REITs?

    ReplyDelete
    Replies
    1. The main purpose of REITs is to create a platform for developer and asset owner to unlock the market value of their assets and recycle their capital. In addition, they also stand to gain from the lucrative recurring management fee which requires low capital outlay. These assets produce stable income which caters to a specific group of people that are looking for stable income and a lower rate of return. However, it differs from bond in that the unitholders stand to benefit from any increase in net property income coming from rental reversion or AEI.

      That is not to say that REITs are unable to add value. If you have observed the retail malls, those owned by REITs are likely to undergo AEI and offers a better shopping experience to those not under REITs. Developers are more concerned with the higher margin development project than to be concerned with enhancing the current investment properties. CMT bought iluma, enhanced it and Bugis + offers a much better shopping experience and has been able to attract stronger shopper traffic.

      The main problem with REITs often lies with the incentive of the management and the stance of the sponsor or major shareholder. Some are perpetually keen on acquisition deals and often acquire assets that are not truly yield accretive. To some sponsor, REITs is an ATM where there are able to push through deals regardless of the opposition of the unitholders. Many have also learnt the lesson from GFC 2008 on the importance of not being over-leveraged and to have a diverse source of funding and well-spread debt maturity profile. Sadly, some have still not learnt the lessons.

      I will say whether REITs can create values really depends on the management team involved. The track record often can show who are the better ones and the market seems to have been rather efficient as seen from the differing dividend yield that REITs in similar sector trades at.

      Lastly, I will say that the current management structure of REITs incentivize the management to acquire more assets given that they often earn base fee as a % of asset regardless of whether there are growth in DPU for unitholders. My personal opinion is that the internal management model adopted by The Link and many other REITs in US is probably a better management structure.

      Delete
  11. Hi there just browse through your portfolio , any reasons why you sell silverlake axis and hour glass ? Just curious as i am vested in silverlake and hourglass under watchlist.

    ReplyDelete
  12. Hi, firstly congratulations on the track record and I really like the work you are doing, you have a good analytical process and look forward to reading more of your posts.
    Just a minor comment on your statement about the difference in value between a freehold vs 99yr property. It may seem ridiculous that the premium is so small but if you actually discount the future NPI into perpetuity (freehold) vs 99yrs, you would realize the difference is really about that amount.

    ReplyDelete