The investment was a good one - shares of HDFC Bank are at Rs. 2,400 today. But HDFC isn't about to dilute it's stake, so the market value is irrelevant; the issue is about how much they pay to service the debt taken to own this stake.

Already, they've paid more than 600 cr. (interest on the whole amount for two years). They've received about Rs. 28.5 per share in dividends, which is around Rs. 67 cr., in the last two years. (For the shares bought only with this borrowing) Net-net, they paid out Rs. 550 cr. .

Of course, they gained, over Rs. 2,000 cr. from the share price increase, so in that context it continues to be a good deal; but the break-even point for those shares has gone up from the initial 1,520 to about 1,800 today.

They'll have to roll this 2,000 cr. over, and the interest rates are likely to be 9-10% this time around. That implies a payout of around 180 cr. on this tranche per year and by 2012, when the next tranche comes due, the total interest paid by HDFC for the purchase will be Rs. 1,000 cr. (or about Rs. 425 per share).

The other thing HDFC had sold then was options at Rs. 275 (300 cr. worth), which allowed investors to buy in at Rs. 3,000 a share. The share's gone through a 1:5 split, so the breakeven price was Rs. 655 for the investors - HDFC trades at Rs. 675 today, though it was lower than that price a few days back. To convert such options, the buyers have to put up Rs. 600 per share (they've already paid the premium) and HDFC will issue them shares. That gives them the money to pay for the HDFC bank shares without needing to borrow much. Will these options get converted?

Currently the answer has got to be "yes", even if the price dips below 655,  because the loss to investors for non-conversion would end up being Rs. 55 per share. Any price above 600 almost surely means conversion. But there is some risk - dilution reduces the price of shares, for one, although at 3.5% dilution that impact is low. And if investors want to realize the profit, they must pay HDFC the money and sell the shares immediately, since nearly 3000 cr. worth is the purchase value. Such selling pressure can't really be sustained (HDFC traded 222 cr. in the whole day today).

(To those confused, there are two entities: HDFC and HDFC Bank. The former owns about 22% of the latter)

As August nears, the action will unfold. Holler if you hear something interesting.

Going pretty well, this story is.

Reasonably revenue growth, nearly 20%. Interest payments went up lesser. This is non-consolidated – last year, their consolidated profits had an EPS of 15% higher (110 cons. versus 95).

Only pressure going ahead is the increase in interest rates, which hurts the company (max cost is interest payments). HDFC’s subsidiary, Gruh Finance had commercial paper going at 9.45% for 68 days – the liquidity piece is very tight. Additionally, a few bonds come due this year and will have to be rolled over. They have increased their lending rates (by 1.25%), which isn’t so high it will start causing defaults. But there is a breaking point, perhaps in the 13-14% range (most loans now are in the 11% range). Finally, there is a NHB regulation that teaser loans will need 2% provisioning – and teasers are 27% of HDFC’s book, which will hurt them going forward.

Haven’t ever liked the stock, as it used to be overpriced. But it’s shown a stellar set of numbers today. Got battered 4% though, as everything financial went for a toss today.

This post is written by Deepak Shenoy, at Capital Mind.

HDFC increased its Retail Prime Lending Rate (RPLR) to 15% yesterday – see pic below - , up 75 basis points (0.75%) from December 1. This follows an increase in RPL increase of 0.5% on September 1.

(Excuse the horrific graphic skills)

They’ve also stopped their teaser home loans – offered at 8.5% until March 2011, and 9.5% till March 2012, with a floating rate thereafter. Typical “thereafter” rates were RPLR minus 5% (4.75% to 4.5% advertised, but many have negotiated to 5%)

That means if interest rates stay where they are, home loan buyers will end up paying 10% on their home loans post 2012, and further increases till then will hurt them more.

To give you an idea, imagine a Rs. 50 lakh loan taken in October 2010.

The 8.5% teaser rate would involve an EMI of Rs. 43,400 till March 2011.

If you took the loan in October, you’d have paid 6 EMIs till March 2011. That leaves a balance of 49.5 lakhs, which takes the EMI up to Rs. 46,500 in 2011.

In 2012, the rate resets to, say, 10%. That takes the EMI up to 48,100 – with another 18.5 years remaining. That means EMIs go up 1.1x from current. Manageable, one might think – a 10% hike is less than inflation, currently.

But if the rate went up to 12% effective (RPLR went to 17%, say) – the EMI will be 54,500, which is 125% of the current payment. Given that there is no restriction on RPLR – HDFC can set what it wants – there is no reason to assume that rates will remain in sane boundaries. As RBI raises rates, money gets scarce, and the cost of borrowing of an NBFC like HDFC goes up, and they will raise rates accordingly.

You want to know how expensive money is? RBI conducted an auction today for T-Bills – the safest form of debt, since it’s government guaranteed. The 91 day T-Bill sold for 6.94% and the 364 day for 7.27% (all annualized, btw). That’s what the government will pay to borrow, for less than a year. And the 10 year T-Bond rate is 8.11%, and the 30 year, 8.43%. With the spread that small, we’re speaking of a very flat yield curve, and the rule of thumb is that financial institutions make loads of money on a steep yield curve (they borrow short term and lend long term) but not so much when the curve is flat.

It seems 27% of HDFC’s loans are in the teaser rate scheme. That’s more than 25,000 cr.  - which means if interest rates stay high, the reset rates will hit these consumers hard.

Since they’ve stopped teaser loans, all new floaters will likely be at 10% or more. How much that will help offtake is something time will tell – but it is likely to hurt sales. Even then, I think the stopping of teaser loans is good – there is some level of systemic risk such loans introduce.

Disclosure: No holding in HDFC.

This post is written by Deepak Shenoy, at Capital Mind.

• Revenues are stagnant for two years now. Around 2,700 to 2,800 crores, this is disturbingly static.
• Even The Trailing Twelve Month EPS (TTM EPS) has hung around the 80s for the last two years, and zoomed up to 100.3 on the back of a great march quarter – where the blip seems to be in a lower interest cost (1550 cr. versus average of 1700 cr in prior and next quarter)
• TTM EPS has grown a miserable compounded 6% in the last two years.
• With interest rates going up will the cost of funds for HDFC go up? It most definitely will, but then so should revenue because most of their loans are floating rate. But they are providing loans at fixed rate for two years (at 8.25% till 2011, and then 9.25% for a year). Now it’s about managing the spread – if rates go up too fast, the spread will contract. They’ve borrowed 73,000 cr. so if they pay even 50 bps more in interest that’s a 350 cr. hit on profit, about 15% of their annual profit.
• HDFC’s annual report states that the total loan book was around 98,000 crores, and their total borrowings were 95,500 crores. (They have investments as well).

My thoughts:

• I’ve always maintained this share is overpriced but it seems to be a stock market darling. Have largely lost money going short this stock – though I’ve made some as well, I would not recommend going short without deep pockets.
• At Rs. 3040, the P/E is 30. EPS hasn’t grown at even close to 30% for the last 8 quarters. If the March 09 figure turns out to be a blip (the EPS was higher by 9 rupees on a freaky lower interest component) they will go back to the 80s on 12 month EPS, which takes the P/E even higher.
• The insurance and Asset management subsidiaries will be in some trouble as business plans change, and the regulators efforts to cut commissions will seriously hit bottomlines in the short term. Now I’m very fond of HDFCs AMC service, but they must make their businesses far more efficient if they really want to stay in the game. That will require some short term loss-making and a culture change, of sorts.
• We don’t know the metro/large city versus smaller city mix. Current prices seem insanely high in the large cities, and I’m not sure how they’ll continue to grow at the same pace. A drop in growth or a reversal of home prices will not just stunt HDFC’s growth, it’s likely to increase their non-performing loans.
• The upside is that things continue to grow, house prices continue to rise, salaries keep going up and more and more people start borrowing for mortgages. I see some of this happening in smaller towns.
• On a technical note the chart has serious strength – it’s near or at all time highs.

This post is written by Deepak Shenoy, at Capital Mind.

What is the trend in real estate development across cities?

It’s all going suburban in the bigger cities. I will give you the example of Mumbai. Guess where HDFC’s business is mainly from? It’s from the far-flung suburbs of Dombivili, Thane, Virar, Kalyan, areas that far from Mumbai. We have nothing this (indicating the city) side, nothing in Bandra, nothing in Khar. Even in New Delhi, if you see most of the business is coming from Greater Noida, Ghaziabad and Faridabad, which are on the outskirts. Noida has become expensive, Gurgaon is expensive so we get only a handful of applications these days from Gurgaon. Growth is all from the outskirts of big cities. But in the smaller cities we are very active and here prices haven’t gone up that much. For example, prices in Kolkata, Lucknow and Jalandhar have not gone up that much. Our average loan size all-India this year is Rs 17 lakhs, which includes Mumbai and Delhi. If you take them out, the average is down to Rs 12 lakhs. In Mumbai, the average loan would be around Rs 25-30 lakhs.

(Read the rest)

Deepak Parekh is a very smart person and his views are quite interesting. Average non-Mumbai/Delhi exposure of 12 lakhs a loan is amazing. Shows me how overvalued these cities are.

HDFC he says has a loan to value of 65% – meaning they finance just 2/3rd of the property, and I imagine the borrowers has further “black” component as equity. That makes for an overall stable market, except for pockets of overvaluation like Gurgaon or Mumbai.

He makes the case for more cities and for spreading out. If you can get fully ready commercial places in Hyderabad near the airport, for rent at Rs. 15 per sq. ft. why stay in Mumbai and pay the Rs. 60 (min) to Rs. 400 on them?

HDFC’s model also looks very interesting – though I still think they are overvalued, with this kind of attitude and conservative mentality they won’t go insolvent. Overall they are still vulnerable to a real estate bust, but that should only help get valuation to a better point, rather than bring in concerns of survival.

This post is written by Deepak Shenoy, at Capital Mind.

The stock fixed itself a little bit after falling a lot at result announcement, but has been going down for the last month or so (10% down from the highs)

The first three quarters have produced an EPS of about Rs. 65, and the FY09 EPS was about Rs. 80. They should be on to make an EPS of about Rs. 90-95 for the full year, which is about 10-20% growth.

The current P/E, at an is about 25-30 then; which I’d reckon is still high (their EPS in FY 2008 was a whopping Rs. 96, so getting back to Rs. 95 is no big deal).

HDFC benefited from an early teaser rate program of 8%-8.5% for three years, and then a reset. But now all banks have caught on, and some are offering one year teasers as low as 5.5%. And the RBI has been cutting off access to funds for NBFCs slowly, so they don’t get to work the spreads as low as they used to; an increase in short term rates will affect margins adversely. 

Their NPAs are at 0.94%, of which about 4/5th are covered by provisions. NPAs are 90 day delinquent loans; if the real estate market improves this should only come down (or be recoverable) On a book basis, HDFC at 4.2 P/B is one of the most expensive (consider that most banks are available for near 1).

Insiders have continuously been selling. Some seem to be buying but that’s funny accounting – they exercise ESOPs and buy back some shares with a small portion of the proceeds, and show it as a net buy (this is a wrong way to report it, but nobody gives a damn).

HDFC owns stake in their mutual fund and insurance businesses; the insurance company still makes losses, but the AMC fund business generated about 129 cr. of profit last year (margins of 29%). Plus, they have a stake in HDFC Bank which has some value, though they are unlikely to sell it.

Overall it’s the size effect, pretty solid management PR and the subsidiary potential that gives them a slight advantage. But as one of the most expensive fin businesses, it’s not showing the growth of the strong and highly valued players in that space – a better bet would be Axis Bank or even its owned HDFC Bank, both growing EPS in excess of 30% yoy.

This post is written by Deepak Shenoy, at Capital Mind.

These are zero coupon debentures - meaning you don't get paid during the tenure. In this case you get a lumpsum principal plus "premium" at the end. The premium is such that your yield is at the rates mentioned. HDFC will really be paying out Rs. 296 cr. for the 2 year debenture and Rs. 508 cr. for the 3 year debenture as total interest.

From an interest outgo basiss, that's about 800 cr. in three years; 317 cr. for the first two years, and 170 in the third. They need to make at least that much to pay for the loan.

In addition, HDFC is issuing warrants - that give you the right to buy an HDFC share at Rs. 3,000 anytime in the next three years, for an upfront non-refundable payment of Rs. 275. This sounds like an American stock option, and it is. So can it be valued that way? Using the Black Scholes model, a three year term with a 3000 "strike" price and a current market price of Rs. 2440, I get a valuation of Rs. 375. (Volatility of 25% assumed)

(Note: Black Scholes is a notoriously unreliable model for valuing long term options - it either overstates or understates prices dramatically. The implicit assumptions of volatility and distribution are totally wrong as observed. The basic underlying assumption of Black-Scholes is that markets are efficient; any observer will tell you that is bunkus. Specifically, you might see "efficiency" work in the short term, but over a long term the model simply doesn't apply)

Still, HDFC gets 300 cr. it doesn't have to pay back, for issuing these warrants. They'll dilute just 3.5% - consider that there is a 4.5% dilution with the total outstanding stock options remaining and another 1% from FCCB conversion. (Heck, that's a lot of dilution - but different story).

They will use the money to buy out 3% of stake in HDFC Bank. HDFC had paid 400 cr. as an advance last year for buying 4000 cr. worth stake at Rs. 1520 per share - this is valid till Dec 2, 2010. Current HDFC price is at 1480 - but if they choose not to buy shares they lose the 400 cr. advance. So they have to pay out Rs. 3600 cr. to say hello to the 3% extra share in HDFC Bank.

While this is good for HDFC bank, the share of HDFC in it goes up from 19 to 22%. That's not much really - a rough calculation shows that even if HDFC Bank grew 20% a year, HDFC's extra stake will only give it Rs. 80-90 cr. a year as increased profit share.

HDFC benefits from the warrants it sells - the 300 cr. it receives can be used to pay for the first year's interest on the debenture. Since it's borrowing to buy HDFC bank shares, it can't really generate cash from that avenue (and it plans to NOT sell HDFC bank shares at all) - so where do they get the money to return this loan after two/three years? I don't know. Maybe they'll borrow again. Scary strategy, that. What if the lenders say F.O.?

Now the extra interest outgo is about 30-40 bps (0.3% to 0.4%). (Of a total borrowing of 80,000 cr. the extra interest is 300 cr.) That may not sounds like a large amount, but they had a net interest margin (inflow minus outgo) of only 2.1% or 210 bps. This extra 30 bps will hurt, from 2011 onwards.

But their prospects are good, no doubt? Well, they have a higher cost of funds than, say, banks. Their insurance and mutual fund subsidiaries are not going to do well in the next few years, considering new tax laws, lowered commissions to agents and higher competition. The new tax law may hurt housing too - tax exemptions are due to go away in 2011, and that will hurt prices. (Housing is where HDFC makes its core money) There is likely to be a revision of NHB provisioning too - current levels say 90 day delinquent loans need to provision just 10%, upto 15 months of delinquency (!!!). Even at 50 months or more of non-payment, the provisioning needs to only be 50%. These rates will change once loans go "underwater" - a situation we haven't quite seen yet, but I believe we've overbuilt enough to see it in the next few years.

The other point is price. HDFC is valued at more than 30x past earnings (it had an EPS of 78 in FY 09, and an EPS of Rs. 19 in Q1 FY10). It's price to book is close to 4 - while most regular banks are quoting at close to book values (p-to-b of 1 or less). It's dramatically overpriced. (It's a good trading stock though - has decent volatility)

You and I can't subscribe, so I won't even talk about what I would do. Still, one has to have some direction. At a 7.5% yield the debentures look pricey - why would I go there, when L&T Finance or a Tata Capital gives me a 10% yield? And the warrants - if I think the equity is overpriced the warrants are a joke. But there are enough takers for the issue, apparently. When the music is playing you have to dance, no?

This post is written by Deepak Shenoy, at The Indian Investor's Blog.