topofeaturellc

I don't think its totally fees with VC's - its just a horrible asset class. Even if you took fees out returns are still bad. The other thing that's amazing is that the reason why the name brand guys outperform if you look at the data is that they get offered a cheaper price to invest at. So it ends up being something like : coin comes up heads a few times, now guaranteed even if you are a market performer to outperform your peers by like 20%. Its crazy. I just had to stop talking about business with my friends who are in VC funded bizzes.

I am under the perhaps mistake impression that a modern CCGT is flexible enough for most standby power - especially when its for load rather than balancing - so its not quite as bad as needing aeroderivatives.

I may be reading that wrong, but that price is inclusive of grid costs. The problem is that grid is always an ROE remunerated business even if its priced on units. So if people take less volumes off the grid, but remain connected to it, they'll raise the price per unit of the grid costs. So in the short term, yeah it only needs to beat the fully loaded cost, but in short order it will really just be pricing against generation, and eventually it'll be generation + the costs of maintaining a less stable grid. Eventually solar will work against all that, but its not as simple as looking at a your cost per Kw delivered and saying that's the bogey SCTY has to hit.

no it won't - because its a commodity and everyones costs will come down the same.

you need to figure out how much capital he needs to invest to supply that much energy. Also you need to figure out what the marginal supplier of electricity will be under your preferred solar scenario, and then what that implies for his return on capital. If Solar is marginal when its producing (and I'm guessing it will be) its over the long-term a cost of capital on marginal capital business - except that the capital cost/MW will probably be declining the way semi costs have declined, so return on the back book will suck. So in reality it'lll be worth less than book. None of this will matter though as its an awesome story that they'll pitch as an EBITDA ramp story, the stock will absolute smoke, and schmucks like me will short it until we're out of business.

Well that's not quite what he said. He said he would be happy using it, but he things it has one too many factors - i.e. the ROIC calc detracts from performance. You can test it on portfolio123.com (with Compustat data) back to 1999 – it's not that far back but as far as it goes I don't see that he's got a valid point there. At least in my test (1-year rolling periods), the Magic Formula out-performed pure EV/EBITDA or EV/EBIT by about 6% per year. His is hardly the only analysis I've seen, and the data Greenblatt himself talks about is supportive of the statement. I've played around with something like this (as well as other quality factors) and arrived at a similar conclusion. 6% a year BTW is a massive number to outperform EV/EBIT by. Massive. That' would be like almost 10% annualized over the market return. I would be curious to know what other quality factors you have looked at. I am interested to see how something like cash earnings/low accruals would work. Seems like you just need something to screen out some of the frauds and accounting oddities. To me a one year snapshot of ROIC isn't the greatest factor for that. Your point about the impact of quality on drawdowns is probably pretty insightful. One could probably argue Buffett brought in the quality focus for that reason around the time he started using float and drawdowns could become really problematic for him. Have any of you guys looked much at the ishares enhanced etfs? They combine value and quality ("factor tilts") , but I think they use P/B and cash accruals for their value and quality proxys. The tickers are IELG, IESM, IEIL and IEIS, (I think) for the domestic and foreign etfs, respectively. They all have low expense ratios. Also, Wes Gray (Quantitative Value) is going to launch some etfs it looks like this year. I will be interested to see what the expense ratios are on those. Another interesting thing to me is that the international mf mutual fund was getting smoked by its index and now I notice gotham is only offering domestic funds. Must be some problem with getting good data. Accruals used to be awesome. Its like one of the only factors I can think of that actually got arbitraged away. I think Quantitative Value has some data on it. Also Empirical Value Partners has done some work on it. A lot of smart value L-S guys were using it to drive their short book during the dot.com bust. My work was a little weird in that I was specifically looking for factors that parsed value traps from value opportunites in a universe defined by the cheapest quintile of something like EV/EBITDA-Capex. If you haven't checked it out AQR has a paper out there on something they call the quality minus junk factor that's worth your time.

You would need to figure out if the chinese government was willing to subsidize loss making coal mines and what level of return they would demand for reinvestment in capacity. It might actually be rational for them to keep cash burning mines running if it keeps employment up and insulates them from foreign issues. Not to mention the chinese government has shown time and time again a willingness to accept a crazy low ROI if it promotes employment. I suspect places like Indonesia will come off the supply curve before China - which implies a normal marginal cost lower than the data implies.

you don't use coal fired plants for grid stability. I'm sure you could find a path from nat gas demand = higher price = higher coal prices, but coal fired plants throttle up and down much more slowly than do modern gas fired plants. Look at the case of germany where solar is a huge part of the network. Even though there is a demand for power to fill in the gaps when the renewables aren't available, coal isn't being used to fill it. Coal works best as baseload or at the low end of the mid merit where you can run most of the day.

Yes especially on the price side. Another option to consider if you believe your LPs demand a name brand law firm is to use a non-NY partner at a big national firm.

Well that's not quite what he said. He said he would be happy using it, but he things it has one too many factors - i.e. the ROIC calc detracts from performance. You can test it on portfolio123.com (with Compustat data) back to 1999 – it's not that far back but as far as it goes I don't see that he's got a valid point there. At least in my test (1-year rolling periods), the Magic Formula out-performed pure EV/EBITDA or EV/EBIT by about 6% per year. His is hardly the only analysis I've seen, and the data Greenblatt himself talks about is supportive of the statement. I've played around with something like this (as well as other quality factors) and arrived at a similar conclusion. 6% a year BTW is a massive number to outperform EV/EBIT by. Massive. That' would be like almost 10% annualized over the market return.

There are some guys who come pretty close to pure value fundamental quant, but they aren't retail, and are surprisingly expensive.

The research I've seen that ROIC detracts from performance but reduces drawdown risk. that makes sense to me intuitively

Well that's not quite what he said. He said he would be happy using it, but he things it has one too many factors - i.e. the ROIC calc detracts from performance.

In IAS total equity does include the minority interest so it needs to be backed out. US GAAP you need to check - but while it is an option for firms to include it in equity, mostly they include it in liabilities - so there is no adjustment required.

Because the Balance Sheet should be a picture at any given time of what we use (assets), of what we owe (liabilities), and of what we own (equity). This is the reason why IFRS require securities to be marked to market. Instead, land, buildings and equipment should be recorded at… what?? If my apartment were on the Balance Sheet of my company as an asset, initially recorded at cost, later depreciated over the course of 40 years, adding back all the maintenance capex required, it would still be recorded at a value little changed from its initial cost… At best its recorded value would have kept up with inflation, if maintenance capex were higher than depreciation charges. But what about the true economics of the city, the neighborhood, even the street where my apartment is located? What if Milan’s economy booms, what if a new subway is built not far from where I live, what if some very fashionable cafes are opened nearby? I am positive during the last 30 years the value of my apartment has appreciated much faster than inflation. If such an appreciation in value is not reflected on the Balance Sheet, that document ceases to be a reliable picture of what I truly own today. Gio So then a house in the outer ring Phoenix exurbs in 2007 marked to market was properly valued? Neither solution is perfect, but in the case of real estate where historic returns in aggregate approximate inflation, historic cost isn't terrible, and is unlikely to overvalue the assets. Don't you see the inherent contradiction between calling for marking assets to market on a Balance Sheet and being a value investor who does not believe in efficient markets? It introduces the worst kind of pro-cyclicality to the accounts.

Why would you assume market value is correct? Or that liquidation could occur at current market? And if you do why don't you just index?

Its a question of semantics and accounting rules. From an economic perspective if something is expensed as incurred or capitalized and depreciated is sort of irrelevant (Tax aside). The point remains buildings deteriorate over time and require reinvestment to stay in shape.

Maybe... But, if I decide to sell my apartment, I will find someone tomorrow morning willing to give 10 times the cost of the underlying land... Because I would not be selling square meters of land, but square meters of an apartment... An apartment in a building that, if depreciated over a 40 years time horizon, would be worth less than zero today! Cheers, Gio No it wouldn't be carried at zero because you pay to maintain it.

Yes, but you and others are missing the points that 1) the maintenance is already accounted for. 2) There is no amount of maintenance you can do on any vehicle or manufacturing equipment which will let you use those things forever. They will need to be replaced in a predictable time-frame regardless of what you spend on them. This isn't the case for buildings. With a reasonable amount of capex (including insurance against disasters) they can last and hold their value indefinitely. 1) Huh? You depreciate and add the maintenance capex to the PP&E 2) I've been in a lot of factories. I assure you that this is not absolutely true. I remember being in substation once where a decent amount of the equipment had originally been installed 70 years ago, had a depreciable life of 40, and Mgmt felt like with appropriate capex would never need to be out right replaced.

there is actually a significant amount of academic research that confirms the idea that it is indeed the land that is appreciating in value while the building itself is a slowly depreciating asset. As someone said above that maintenance capex is slowly building "apartment 2.0"

I'm with Geo on this one. Even with proper care and maintenance most equipment doesn't hold its value and will eventually need to be replaced. Equipment either deteriorates in value to the point that it isn't worth fixing it when it breaks or it becomes obsolete and it is worth the investment to buy newer equipment. In other words it depreciates. None of this is true with a properly maintained building. It holds its value quite well if maintained properly and doesn't become obsolete by the newest iBuilding 2.0. Well that's why you capitalize certain costs of maintaining the building and why depreciation schedules for buildings are 30-40 years.

Its plausible that $1 of incremental capex is worth less than one dollar of EV.

I tend to think PP&E/Sales is a better metric to focus on than absolute capex level. I also find that most managements definitions of maintenance vs expansion capex is closer to "keep the machines running" as opposed to "adequately support organic growth" So in the context of your model - I assume capital intensity approximates that - and yet it goes from 11% ish, to 16% today, to 18% - and your marginal capital intensity is even higher. So if you like those assumptions, then its totally plausible there is no equity value on their current business model, or somewhere better 11 and 16 is sustainable and you've understated their ability to generate cash.

Well...you've sort of figured out a key insight. Growth alone tells you nothing - its the capital required to generate that growth and the return on said capital. think of ROIC in a kind of dupont framework where you have NOPLAT margin * sales/IC. The more capital intense a business is the more important Sales/IC becomes as a driver of intrinsic value. For a service business with no capital its hardly worth thinking about capital intensity or even ROIC on the base business - instead focus on margin and where the excess cashflow goes.

why don't you save it as a google doc or dropbox and post the link here.