Investing

With all the talk of interest rates I was considering home buyers and how they could hedge against rising rates if they were not ready to buy today. Since working on some of the numbers last night Motley Fool had an interesting article on home owners and refinancing. It turns out that nearly 25% of all home loans originated in 2002 had some variable rate provision such as an ARM (where the rate is fixed for 5 years and then floats). Those ARMS would have $300 more a month to pay when their rates go up by 40 basis points in five years based on a 100k house. How financially illiterate must people be to not lock in historically low rates. You can get a 30 year mortgage for around 9 points more than the same 30 year Treasury bond. That is about 5.8% for the mortgage with 30 yr treasuries at 4.9%. The historical mean from 1972 to 2002 for mortgage rates is 9.7% (median 9.05%). The standard deviation is only 26 points; we are 2 standard deviations below the mean right now for 30 year mortgages!!



So what are some of the hedging strategies available? Well unfortunately you can't just lock in the loan today and buy your house in 2.5 years time; this would be the simplest solution. Really all you can do is calculate the additional cost if mortgages go back to historical levels and then try to make this money through some hedge. From today's 5.8% rate a rise of 10 basis points adds about $12k to the cost of your loan for a 100k house (Discounted back to today. That is this considers inflation and the fact that you don't have to pay the money until some time in the future and you could invest it today). Each 10 basis point rise adds about another $12-$13k. For each drop of 10 basis points your cost reduces by $10-$11k.

I spent a while investigating what financial instruments could be used to hedge this rise. The only one I could find were the CBOT treasury options (mentioned in this post). There are options on the yield of the 30 year Treasury bond, these are cash settled at $100 * yield change for each contract. As we expect rates to rise and we want to profit from that rise then buying calls would be the most appropriate strategy. A hedge using only calls costs about $11,025 for a 2.5 year option (CBOT apparently have TYX-30 year treasury LEAPS, but I couldn't find their pricing so I used Black Scholes based of the implied volatility of the shorter term options listed).

* If yields rise by 10 points your option would generate $3150 in cash but cost you 11k to enter and your mortgage now costs $12k more. Net you are out $20k.
* If, however, yields rise to 7.8% your option generates $24k in cash, cost you 11k to enter and the mortgage is $24k more. Net you are out $11k.
* If yields rise to 8.8% your option generates $45k in cash, cost you 11k to enter and the mortgage is $38k more. Net you are out $4k.

You could offset this somewhat by selling puts and effectively entering a collar. The put prices are very low though, as the 30 year bonds are not likely to go much lower. Selling puts could decrease the call price by about $2400 and if rates drop 10 points you owe about $11k on the puts (not a great position to be in when you are about to apply for a mortgage) immediately, even though you have saved $11k in today's dollars due to the rate decrease. The difference is that you have to pay the $11k today to settle the put, the savings in your mortgage come over the next 30 years. The chances of rates decreasing to 4.8% for mortgages in 2.5 years seem quite remote. Importantly the TYX options are European style, (that is they cannot be exercised until expiry day) so there is no risk in the intervening time.

Given that rates are likely to return to around the 8.8% level by the end of 2005 your 11k on calls saving you 45k in mortgages is probably well spent.

Interestingly the market somewhat hedges you automatically. When interest rates are so low, housing prices increase dramatically so a $100k house today may sell for closer to 60-70k when interest rates increase back towards the mean of 9.7% or even higher. By waiting 2.5 years to buy a house you could be hedging your interest rate exposure today and taking advantage of historically low rates. Additionally by waiting for a few years you can also take advantage of lower housing costs when money is more expensive and those with 5 year ARMS have $300 more a month to pay.

CVCO is down to 18.15, having lost 9.25% today. I hope it keeps falling. It seems likely that the S&P 500 funds are still selling after all with such a large drop over the last two trading days.

A few notes on the treasuries analysis from yesterday. Firstly a move in the yield isn't directly proportional to the bond price. The bond price is a function of the time to maturity and the yield. The 30 year bonds currently have a duration of 14.7 years, that means for every 1 percentage point increase in yield, the price will drop by 14.7%. Therefore from the current 4.5% yield to the mean of 8.3% you would expect a return of 55.86%.

The Rising Rates Pro Fund actually tracks the inverse plus 25%. That would take the return to 70%. I had indicated I thought a double was likely, I would revise that to 70% in a few years as a sure thing. A 100% return could still easily happen if the yield rose to 9.94% (though 77% of months were below this).

I have been thinking about shorting treasuries for a year or so. It turns out to be really complicated so I haven’t actually done anything yet. While watching one of my many finance programs this weekend I found out about a fund whose objective is to increase in value by 125% of the decrease in the value of a 30-year treasury bond. The Rising Rates Opportunity Pro Fund requires a 15k minimum investment and has a management fee of 1.94% of assets annually (which is only a little higher than the 1.85% fee on their S&P 500 index, compared to the Vanguard S&P 500 index fee of 0.18%!). Though the fee seems very high, it isn’t particularly high for a managed fund and is probably much cheaper than the alternative ways of profiting from falling interest rates.

I had considered Chicago Board of Trade cash settled interest rate options. The problem with them are threefold, expensive to enter, you are paying a large premium for the time value of the option and the longest time that you can purchase the option for is 12 months out. Combining the Rising Rates Pro Fund with a put at 45. That would mean for every option I hold I would receive $100*(45-(new yield*100)). That is a yield of 3.5% (totally unheard of) would pay $1000 per $170 worth of puts to make up for my losses in the fund (this would be a hedge). The problem with options is I am not just asserting that yields will rise but I am also "betting" that it will happen within 12 months.

Other alternatives include interest rate futures or shorting one of Barclay’s Treasury Fixed Income iShares such as the 20+ Year Treasury Bonds (TLT).

Before deciding on my course of action I wanted to find out just how historically low the long term treasuries were. I downloaded 30 year bond data from http://www.forecasts.org/interest-rate/ . They have data back to Feb 1977. Looking strictly at this data the 30 year bond median is 8.3% and the median is 7.9%. I didn’t feel that this data really went far enough back and additionally the fed adjusts their discount rate rather than the actual 30 year bond yield. I looked at the correlation of fed funds rates to the 30 year bond yield. The correlation was 0.85 (where a correlation of 1 would mean they move completely in harmony and 0 would mean there is no relationship). Given such a strong correlation I looked at the fed funds rate from 1954 to now as a proxy. The mean difference between the 30 year bond and the fed funds rate was 1.19 with a standard deviation of 2 (showing that the 30 year bonds generally do not hit the peaks or troughs of the fed funds rate but track closely around the middle ground). The average fed funds rate from 1954 to today would be 5.92% (the current yield is 1.25%).

With the current 30 year yield at 4.68%, there are only 3 months of data out of 318 where the yield was at or below 4.68% since February 1977. Even at 5% there were only 11 months out of 318.

If it takes 1 year to get to 5% you would have made 6.8%. If it takes 2 years to get back to the mean of the 30 year bond rate, you would have made 77.35%. This is incredibly low risk as even if the fed eases one more time, 30 year bond rates are not going to drop commensurately and eventually they will put rates back up and you will make money. If I was ever going to make a 2-3 year call, you will at least double your money with this investment.

Well CVCO has traded much higher than I had hoped (I don't want to say expected as I'm not sure I had any basis for expectations).

There were a few notable announcements this week that affect the manufactured housing business (and therefore my estimates of Intrinsic Value). These have been outlined by Orbis in an effort to discourage Clayton homes (CMH) sale to Berkshire Hathaway (BRKa). Orbis, a CMH shareholder, sent a letter to other CMH shareholders encouraging them not to sell at BRKa's bid. Some of the interesting points:


1. The outlook for the manufactured housing industry is improving, coming off a cyclical trough

2. The bottom of the cycle is the wrong time to be selling a great company in a highly cyclical industry. We believe in the long-term fundamentals of the manufactured housing industry and that Clayton is an outstanding company with a great future.

3. U.S. Bancorp, the eighth-largest financial services company in the US, announced on June 27 that it would begin offering financing for manufactured housing. Commenting on this news, a leading industry analyst said: "Signs that new ... lender entrants are imminent should mute concern about weak expected near-term results and allow the market to look with renewed confidence toward a more prospective 2004 and 25%+ [industry] shipment growth."(2)

4. In Texas, the largest market for manufactured housing, a new law became effective on June 18 reversing legislation which had severely depressed manufactured housing shipments.

5. Indications are that the overhang of repossessed homes, which crowd out new shipments, will evaporate in 2004.(2)

(2) BB&T Capital Markets report, 6/30/03


Items 3 & 5 are incredibly significant and fundamentally change my likely valuation case for CVCO to the upside. I mentioned in my June 26 post that I believed lenders would start to return in the long term, I certainly didn’t imagine there would be an announcement within a week!

CAVCO traded between 15 and 22.50 this week, closing at $20, down 7% off Wednesday’s close (Friday was July 4th). From what I can make out, about 620k shares have been traded.

I wanted to know how many shares of CTX were held by S&P 500 funds to establish if the sell off is probably over or just beginning – don’t try this at home.


I managed to dig it up after an hour or so. Rough math:
1. S&P Press Release on total S&P 500 Assets for 2002 - 900bn in S&P 500 assets at end of ‘02

2. S&P Growth since the end of 2002– almost 10% growth in S&P 500 this year

3. CTX percentage weight in the S&P 500 market cap weighted index is 0.0527% – (this was horrible, I used the top 10 holdings off the Vanguard 500 website and calculated their total market cap - $2,139,122.29. Then I divided that by the 23.1% of the index that they represent. Result is that the total index is worth $9,260,269.65 bn. Centex Market cap is 4884.21 MM, divided by total market cap gives 0.0527%).


End result – about 6.6M shares of CTX are held by S&P 500 index funds. That is 330,000 shares of CAVCO are held by S&P 500 index funds. Most of them will need to sell relatively soon. Given that 600k shares have already traded, about 20% of the company may have already changed hands. It is possible that the S&P funds are already out. I was hoping that the math would show that there was more selling to come, unfortunately not.

If in fact most of the selling is over, then I am certainly curious as to who (probably a few have taken the majority) has taken this 20% stake in CAVCO. I’m going to watch it over the next week and see how it trades. Friday was a low volume day at only 100k shares, compared with 300 and 200k the days before. It is strange to see such a large drop in price on such low volume; if the funds were more or less done selling by Friday then the abating selling pressure should have driven prices up. I really can't make any conclusions based on funds flow here. I suspect I will be lucky to see sub $10 but I may try wait for $15. With these positive movements in the industry my $31 estimate only assumes 5% growth from these very depressed levels. A 50% discount to intrinsic value is certainly a wide enough margin of safety. Another interesting strategy may be to dollar cost average into CVCO over 4-8 weeks.