No new stimulus, a bit of a market retracement in USDJPY and Japanese equity indices. So far so normal, it also means I don’t have to break out the abacus and do some new technical analysis on the things. Yay for laziness! Where can we find some value then?
Bloomberg is pretty interested in finding better yielding products for Japanese investors, both in debt and in equity. SMFG is doing something pretty interesting here, using its Big Bank status to pool capital and distribute earnings potential in global infrastructure projects while opening up new markets for Japanese smaller-scale funds. Meanwhile, Societe Generale is involved in some pretty (mathematically) interesting index building to figure out ways to uncover stock growth in Japan.
Infrastructure loans pool:
Let’s start with the relatively short and sweet here in the debt department: Sumitomo Mitsui Financial Group is buying in to international infrastructure project debt (unclear if this has already happened to be honest) and selling parts of the portfolio to regional banks in Japan, in chunks of 2-3 billion yen per bank. The interesting thing is SMFG will keep relatively large long positions in the infrastructure debt pool (again unclear exactly how big) reducing interest conflicts between them and the client banks. The question here is obviously how the fee structure is done, but that is apparently an area of negotiation for the parties involved. A couple of interesting things:
- This is literally what I think international capital markets are for! Connecting savings capital with opportunity on a global scale, leaving the lenders better of for either more or cheaper capital, and giving the savers better yield. Win-win-win!
- Given the structure and significant originator shareholdings of this arrangement, why isn’t SMFG securitizing and tranching this product? Come on, there must be billions of US$ in demand for this in Japan alone, from everything ranging from pension funds to Mrs. Watanabe cash savers, corporate treasurers and average brokers. Hell, if there is ever a bunch of good investment opportunities but the Japanese banks are taking too long to provide capital, drop me an email and I’ll run around Taipei getting you a few tens of millions of US$ to get going!
- Using the infrastructure angle is both very interesting as it pretty directly taps into growth and seems very safe since I think the projects selected for this are a little bigger than filling in potholes on random roads. The likely size of the needed funds for a diversified portfolio is really interesting, as is the very probable government involvement in the funding, and as infrastructure is primarily being developed in growth markets it covers that angle too. Exciting stuff!
SMFG, if you ever need a person to market your products in AXJ, or someone to do initial deal prospecting… ah well, one can dream.
Index analysis:
Societe Generale is at it again, doing cool quantitative finance stuff as per usual. This time they’re set to fight the JPX Nikkei 400 Index for the shareholder-friendliness index price. Let’s dig into some technicalities and see where the JPX Nikkei 400 leaves some room for improvement, and what the Solactive Japanese Buyback Index by SocGen is doing instead and how I would compare them. First, JPXN 400 component weights:
- 40% – Return on Equity (three-year simple moving average)
- 40% – Historical cumulative Operating Profit
- 20% – Market Capitalization
So the gauge primarily identifies companies with short-term ROE, stabilizes the volatility that an ROE measure will easily have by applying a simple moving average, as well as checking for operating profits since inception of the firms. Good ideas all, rounded out by filling in the remaining 20% with the firm market cap to give a slight leg up to larger companies to be more representative of the Japanese economy. Most of this was criticized since a lot of western analysts don’t like large Japanese companies (“&%¤# cash reserves!”) and some thought that ROE was a bit too volatile and that the index was indeed pro-cyclical, ranking companies well right before they are more likely to turn down rather than up. (Honestly, for that, maybe you would want to look into companies weighted on capital raised over assets in the last twelve months and percent drop in cost of equity in the same period?)
The SocGen index instead follows this algorithm:
- Screen companies for announced buybacks in the last 2 months.
- Rank them on the percentage of shares outstanding being bought back.
- If:
- Buyback Announcements > 24: Select top 25 buyback ratio stocks.
- Elseif 26 > Buyback Announcements > 14: Select all Buyback Announcement stocks.
- Elseif Buyback Announcements < 15: Select all Buyback Announcement stocks + [15 – BA at current period] of top buyback ratio stocks in the index during the previous period. Endif.
- Weight stocks based on buyback ratios.
- Repeat every two months.
It is pretty brilliant: it rides momentum in relatively direct ROE increases and companies showing a certain degree of confidence, and refreshes often. I think this is very smart – for portfolio construction or as an active manager benchmark! As an index to base investment decisions on? Pardon me for wearing the thickest nerd glasses in existence but not really. Why?
- Inconsistency: How many companies are in the index, are they contributing to buybacks, and when did that happen? I don’t really know, the index won’t show me unless I’m looking at the buyback announcements on a daily basis…
- Leaves out potential for moves: Why not update the index daily if you’re looking at this daily anyway for the news tickers? Changing a bit of code isn’t that hard, and you still have to run the transactions…
- Completely leaves out companies that have a lot of valuable growth opportunities.
- Doesn’t really represent a selection of how well Japanese companies are overall at doing buybacks because of its flexibility and short-term focus.
Again, brilliant benchmark for a fund manager and I’d put beating that benchmark as a prerequisite for paying out bonuses to any active portfolio manager, but I do not expect to see this index quoted anywhere anytime soon.
What would I do here? Well, I’d probably change this index (if we’re so in-love with buybacks) to either a cumulative (or very long-term exponential moving average) gross buyback ratio to make it less volatile, but I still don’t like the idea of excluding companies looking for more share capital to enjoy risky business opportunities.
“Alright, Mr. Nitpick, what should an index you would design from scratch look like?”
Without having done proper large-scale regression testing on all companies in existence, here’s what the principal components of the weighting algorithm would be:
- Cumulative [Operating Cash Flow less Depreciation and Amortization] to Assets ratio. (Is the company generating significant cash flow from operations? Has it been profitable for a long time? Free Cash Flow is not used here since it removes capital expenditures but keeps depreciation and amortization in.)
- Short-term benefits component: Buyback-to-Net Income Ratio + Dividend Payout Ratio, both cumulative of the last three or five years. (Is the company providing short-term shareholder gains? The time period is to smooth over “rough patches”)
- Long-term benefits component: cumulative investments rate, possibly as defined by [CapEx less Depreciation and Amortization] over Assets, summed since inception of the company. (How much has the company invested historically in its future operations? Alternatively, exchanging CapEx with the Net Cash Outflow From Investment Activities in the formula above could be very useful if many companies are investing in acquisitions etc.)
- Capital tie-up term: [(1.5 – Current Ratio)^2 – 2.25] * [Current Assets / Total Assets]. Penalizes companies for tying up more than 1.5 times current liabilities in current assets, and does so more when there are large current assets on the book. Also penalizes companies that have too little working capital as it will negatively impact cash flow in the imminent future.
- In putting everything together, I would probably have put the cumulative cash flow component as a multiple of a regression-tested weighing between the three other terms to form my weighting criterion. This values the company’s long-term ability to earn operating profit very highly, as well as allowing companies to either invest in future prospects or distribute excessive current assets in either buybacks or dividends, while penalizing companies that keep too much cash relatively harshly. In addition, i would probably overweight the short-term component to somewhat neutralize the heavy influence of historical earnings.
I have yet not settled on selection criteria for inclusion into this possible index. (The above is the weighting alone.) An aggressive growth index would check out the proportional asset increases over a period of time, stability-related selection would check for market cap, and aggressive shareholder benefits investigations would select based on capital structure changes or addition of debt in excess of growth in working capital. But I think this is something better suited for someone who could later employ me to decide.
One flaw in this index structure is that it really rakes bondholders and lenders over the coals, but if we’re talking about tailoring an index for Japanese stocks I think there shouldn’t be any cause for alarm!