Sunday, 29 November 2015

Benchmarks for Retail Bond Pricing

There are a couple of retail bonds listed on the SGX. How do you tell whether the 5.25% coupon offered by Aspial's 5-year bond is too high or the 3.65% coupon offered by Frasers Centrepoint's (FCL) 7-year bond is too low? To determine the right pricing for these bonds, you need to know what is the credit rating of the bond and the Yield-to-Maturity (YTM) for bonds with similar ratings. Unfortunately, the majority of retail bonds listed on SGX are not rated. On the other hand, thanks to Bondsupermart, we now have benchmarks to determine the YTM and pricing for bonds with various ratings. The blue line in the figure below shows the YTM for bonds rated A3 by Moody's (equivalent to A- by S&P / Fitch), which is the lowest of all A-grade bonds. The red line shows the YTM for bonds rated Baa3 by Moody's (equivalent to BBB- by S&P / Fitch), which is the lowest of all investment-grade bonds. Also shown in the figure in green is the YTM for the Singapore Government Securities (SGS), which provide a risk-free return for various bond maturities. This green line is also known as the Yield Curve. The difference between a bond's YTM and the yield curve at the same bond maturity is known as the spread and is commonly used to measure the pricing of a bond. Actually, you can construct a different line for each bond rating, but for ease of discussion, we will show only 3 lines in this figure.

YTM for A3/A- and Baa3/BBB- Bonds

The YTM of retail bonds listed on SGX are plotted as purple points in the figure above. The YTM has considered any step-up coupon feature in the bond, which is why CapitaMallsAsia 3.80% bond has a YTM of 4.064% even though its coupon is only 3.80% (this step-up will only apply if the bond does not get redeemed early). Among the 6 retail bonds with maturities of less than 10 years, only CapitaMallTrust's (CMT) 3.08% bond is rated A2 by Moody's. Let us use this as an example to explain how bonds are priced. CMT bond's rating of A2 is 1 grade higher than A3 which is represented by the blue line. As shown in the figure above, CMT bond's YTM is slightly below the blue line. Since CMT bond is less risky than A3-rated bonds, it should have a lower YTM than A3-rated bonds. Thus, CMT's 3.08% bond is correctly priced.

All the other retail bonds do not have any credit rating. We have to estimate a probable credit rating for the bond and from there, determine if it is overpriced or underpriced. Using FCL's 3.65% bond as an example, the YTM is closer to the blue line (A3 / A- bonds) compared to the red line (Baa3 / BBB- bonds) as shown in the figure above. If we think that FCL bond's rating is closer to Baa3 / BBB- than A3 / A-, then its YTM should increase to that of Baa3 / BBB- bonds (i.e. price should decrease). 

How do you compute the price of a bond from its YTM and vice versa? There are simple tools to do so. One such tool is posted on the SGS website. Just enter the maturity date, coupon rate and YTM and compute the price (or vice versa). As an example, the price of FCL's bond is currently $0.990, which translates to a YTM of 3.826%. If we think FCL's bond should have a Baa3 / BBB- rating, its rightful YTM should be 4.216% as shown in the figure above. Based on this YTM, the price for FCL's bond should be $0.968.

Finally, although we do not know the credit rating of Aspial's 5.25% bond, Oxley's 5.00% bond and Perennial's 4.65% bond, they are all priced above the red line, which means that the market thinks they are not investment-grade bonds. This is consistent with the analysis based on Benjamin Graham's criteria of earnings coverage and stock value ratio, which showed that all 3 bonds (together with FCL's 3.65% bond) do not have sufficient margin of safety. Readers will have to assess for themselves if the higher coupons offered by the bonds are sufficient to compensate for the risks in investing in them. 

P.S. I am vested in FCL's 3.65% bond and CapitaMallsAsia's 3.80% bond.

See related blog posts:

Sunday, 22 November 2015

Is Benjamin Graham's Art of Bond Investment Outdated?

There were 4 retail bonds issued this year, yet all 4 were assessed not to have sufficient margin of safety according to Benjamin Graham's criteria of earnings coverage and stock value ratio. Despite so, all 4 bonds have hovered around the par value since listing. Is Benjamin Graham's art of bond analysis as described in The Lost Art of Bond Investment outdated?

To answer this question, let us go back to the first principles. The safety of a bond depends both on the issuer's ability to pay and willingness to pay. The ability to pay in turn depends on the issuer's recurring earning capability and/or adequacy of assets to extinguish the bonds. The willingness to pay depends largely on the management's integrity and is usually safeguarded by convenants in the bond agreement restricting management's rights, such as declaring excessive dividends to shareholders, or allowing bondholders to take over the company in the event of a default. Benjamin Graham's criteria measure the issuer's ability to pay and the 2 criteria of earnings coverage and stock value ratio measure the issuer's recurring earning capability and adequacy of its assets respectively. Although they were derived more than 80 years ago, the concepts are still relevant today and can be found in today's loans, including the housing mortgage loans that most people are familiar with. Interestingly, all the 4 retail bonds were issued by property developers. Let us visit each one of the criteria.

Earnings Coverage

The earnings coverage is more commonly known as interest coverage in today's age. It measures how much cushion the company's recurring earnings has to cover the total interest expense from all loans. The inverse of the earnings coverage would be equivalent to the Total Debt Servicing Ratio (TDSR) introduced by the Monetary Authority of Singapore (MAS) to determine how much of a person's salary could be used to service all the loans he has and hence how much loan he could obtain to purchase a property. Using industrial bonds as an example, Benjamin Graham recommended an earnings coverage of not less than 3 times fixed charges. This translates to a TDSR of 33%. Although this figure looks low compared to MAS' allowable TDSR of 60%, it has to be considered that individuals have mostly stable salaries, whereas companies' earnings are quite variable from year to year. The more variable the earnings are, the lower the TDSR should be. In fact, for bonds of public utilities which have more stable earnings, Benjamin Graham allowed a lower earnings coverage of 1.75 times fixed charges, which translates to a TDSR of 57%, which is quite close to MAS' allowable TSDR of 60%. Considering the variability of earnings of property companies, an earnings coverage of 3 times fixed charges is not excessively stringent. 

Stock Value Ratio

The stock value ratio measures how much cushion the company's equity could provide to the bonds in the event the company loses money. It is measured as the ratio of equity to total debt. In today's age, the related but more commonly used yardstick is the Loan-to-Value (LTV) ratio, which measures the ratio of debt to asset value. Using industrial bonds as the example, Benjamin Graham recommended a stock value ratio of not less than 1. This translates to a LTV of 50%. Again, although this figure looks low considering that individuals could borrow up to 80% of the property price, it has to be considered that individuals pay down their loans over time whereas companies mostly roll-over their loans when they mature. Thus, the LTV for individuals decline over time whereas the LTV for companies mostly stay constant. Moreover, individuals' mortgage loans are secured using the property as collateral, whereas companies' loans may either be secured or unsecured. A more appropriate benchmark would be the 45% leverage limit imposed by MAS on REITs. From this angle, the stock value ratio of 1 for property companies is not excessively stringent. For public utility bonds, Benjamin Graham allowed the stock value ratio to go up to 2 times, which translates to a LTV of 67%.

Other Considerations

Having too much debt does not necessarily means that a company will default on its bonds. It is akin to an individual with high cholesterol; while it places the individual at risk of a heart attack, it does not mean that one will definitely occur. The catalyst for a heart attack for companies would be a credit crunch or a sharp or prolonged industrial downturn in which the company is unable to find a bank willing to refinance its debts or a buyer willing to buy over its assets at reasonable prices. One of the biggest lessons REIT investors learnt from the Global Financial Crisis is that in times of crisis, having a strong parent helps. REITs with deep-pocket parents such as CapitaComm or A-REIT weathered the storm better than REITs without such parents. 

To conclude, Benjamin Graham's art of bond analysis is still as relevant today as it was formulated 80 years ago. 

Finally, you probably have heard the story that banks are famous for being fair-weather friends; they are nowhere to be found when you need the money but beg you to take the money when you do not need it. As bondholders, we should learn from the banks, i.e. lend money only to those who do not need it!

P.S. I am vested in Frasers Centrepoint's 3.65% bond (held as available-for-sale instead of held-to-maturity).

See related blog posts:

Sunday, 15 November 2015

Does FCL's 3.65% Bond Have Sufficient Margin of Safety?

I realised I had omitted the analysis for Frasers Centrepoint's (FCL) 7-year, 3.65% retail bond when it launched its Initial Public Offering in May. Since FCL announced its latest full-year financial results a week ago, here is the analysis according to Benjamin Graham's criteria of average earnings coverage and stock value ratio.

Earnings Coverage

Profit before tax, fair value changes & exceptionals = $955.4M
Adjusted for:
- Deduct: Share of associates' & joint ventures' results = $279.4M
- Add: Finance cost = $186.2M
Total earnings available for covering fixed charges = $862.1M

Finance cost = $186.2M

Earnings Coverage = $862.1M / $186.2M

= 4.63

The earnings coverage of 4.63 times is above the minimum average earnings coverage of 3 times for industrial companies.

Stock Value Ratio

No. of shares = 2,895.0M
Share price = $1.63
Market value of shares = $4,718.9M

Total borrowings = $9,255.3M

Stock value ratio = $4,718.9M / $9,255.3M

= 0.510

The stock value ratio of 0.510 is lower than the minimum stock value ratio of 1 for industrial companies.

Quantitative Assessment

Thus, FCL's bond meets the earnings coverage criterion but not the stock value ratio criterion. Based on Benjamin Graham's criteria, FCL's 7-year, 3.65% bond does not have sufficient margin of safety.

Other Considerations

Although the focus of the financial assessment should be on the company issuing the bond, we could look towards the parent company if the issuer company does not have adequate financial resources. In the case of FCL, it is the subsidiary of F&N, which is in turn a subsidiary of TCC Assets. If we can be sure that F&N will not completely sell off FCL, then we can assume that F&N will come to the rescue of FCL should it encounter difficulties in paying interest and/or principal.

Thus, for the above reason, I am vested in FCL's 3.65% bond even though it does not meet all of Benjamin Graham's criteria (Note: it is held as an available-for-sale instead of held-to-maturity investment).

Sunday, 8 November 2015

No, I Didn't Buy OSIM at 6¢ During GFC

I was originally blogging on the "bear market" theme before the unexpected stock market recovery in early Oct and the retail bond issuances in mid Oct disrupted the plan. Since the "big bad bear" looks like it is not coming any more, let's end this theme with this post.

You might probably have come across blog posts that said if you had bought OSIM at its all-time low during the Global Financial Crisis (GFC), you would have become very rich. That can be quite true. If you look at the figure below, OSIM touched an all-time low of 5¢ before recovering to a high of $2.90, for a 58-bagger. It is the stuff that most investors could only dream of.

OSIM Share Price

But no, I did not buy OSIM during the GFC and here is my side of the story. Please do note that when I say "thinking of buying OSIM" in this post, what it really means is that the thought crossed my mind but I did not seriously intend to put money into it. It had too much debt for my liking and it lost a lot of money during that period due to a poor acquisition in US. I did not put it on my watchlist, hence, I only saw it from time to time. It is like entering the supermarket occasionally and seeing something on sale and thought of buying it but eventually walking away without doing so. Nevertheless, a decision not to do something is as real as a decision to do something.

The first time the thought of buying OSIM crossed my mind was when it was trading at $1. I told myself that I would consider buying it when it reached $0.60. $0.60 came, but I decided to give it a pass, since many stocks were falling then. The next time I saw OSIM, it was trading at 6¢, where it was near its all-time low. If you think that I was rejoicing at the once-in-a-lifetime opportunity to load up heavily on OSIM, you guessed it wrongly. The thought then was "Luckily I didn't buy OSIM at $0.60!". Because if I had bought it at $0.60, I would be facing a 90% loss within barely a year of buying it. 

I did thought of buying it at 6¢, and even considered the execution costs and risks of it, but eventually did not do so. Sometimes, when we imagine buying a particular stock at a certain price, we always assume that we could get all we want at the desired price. The truth can be quite different. At that time, the bid-ask spread was not 0.1¢ as it is now. Then, the bid-ask spread was 0.5¢, i.e. buyer bidding for 5.5¢ and seller asking for 6¢. The spread by itself already constitutes 8% in execution cost should you buy directly from the seller. Not only that, liquidity dried up during that difficult period. My investment (if I had bought) would be around $5,000, or 80,000 shares. I suspect that these 80,000 shares could not be filled with a single order. Some might be filled on the following day or even at 6.5¢ if I was really impatient. At 6.5¢, that is another 8% in execution cost. In total, the execution cost/ risk can go up to a significant 17%. With an investment of $5,000, that can mean between getting 90,000 shares (if bought at 5.5¢) or 76,000 shares (if bought at 6.5¢). It can make a huge difference in the eventual profit that you could make from OSIM. You could see similar trading difficulties in some of the stocks in today's environment, such as MTQ.

Although OSIM is the biggest multi-bagger to be had during GFC, it is by no means the only multi-bagger available. Around the same time as I saw OSIM at 6¢, I was seriously deciding between buying Valuetronics at 7.5¢ or Hongwei at 12.5¢, eventually choosing Hongwei over Valuetronics. Had you bought Valuetronics at 7.5¢ during the GFC and held to its all-time high at $0.585 in Jul last year, you would have reaped a 7-bagger. While it pales in comparison to the 58-bagger of OSIM, it is still a very profitable transaction. In fact, a lot of investors did very well post-GFC despite not having bought OSIM.

Buying (or not buying) OSIM at 6¢ is still not the end of the story. There is still the recovery side of the story. The next time I saw OSIM, it had recovered to $0.23. Did I regret not buying OSIM at 6¢ at that time? Yes, I did. But I probably would have sold most of it at $0.23 had I bought it at 6¢. At $0.23, that is almost a 4-bagger. Considering that OSIM was losing a lot of money then (it lost $0.18 per share in 2008), I would be happy to sell most of it at $0.23, just in case it came back down again. As the price continued to climb, I would probably continue to sell it, with the last lot at $0.60 for a 10-bagger, exactly the price that I initially thought of buying it.

Putting everything in perspective, although not buying OSIM at 6¢ was regrettable, the regret pales significantly in comparison to the pain of selling most of OSIM at $0.23 and watching it climb all the way till $2.90! This is why whenever I recount the events of GFC, I would regret not having invested more in the stock market then, but I have not regretted not buying OSIM at 6¢.

In conclusion, it is easy to say "buy OSIM at 6¢ during GFC". But there are a lot more considerations to executing such a buy decision. Even so, not buying OSIM at 6¢ might not be such a bad thing. Most importantly, you do not need to buy OSIM at 6¢ to be successful in investing.

See related blog posts: