U.S. Bancorp's Greatness Is Priced In [View article]
Thanks for your comment Toeser. There are two components of risk as I see it, there's the inherent business model risk and the risk that you've overpaid for a stock. I fully agree with you that USB's business model has less risk than both C & JPM. I also agree that this difference in business model risk warrants a premium multiple. However, I think that the market already more than reflects that difference. Just by way of comparison, below is the current P/T1C ratios for all three banks:
P/T1C (using Friday's close) C 0.72x JPM 1.04x USB 2.48x
As you can see USB is trading at a substantial premium currently. I'm also not calling for the gap to close entirely. Again, as I believe USB's business model is lower risk and therefore deserves a premium. Below I show my targeted long-term P/T1C for each of these banks:
Long-term Sustainable Target P/T1C C 1.1x JPM 1.4x USB 2.7x
However, when you look at what that implies in terms of price appreciation even if we assume no growth in Tier 1 common from earnings retention, we'd expect the following appreciation in price:
Price Appreciation from Achieving Tgt P/T1C Ratio Immediately C +53% JPM +35% USB +9%
As I see it, the fact that the stock is trading at a discount to my target multiple is part of what's giving me the margin of safety.
One final point, in one respect since I'm deriving my ROT1C by looking over the last ten years and since Citi faired so poorly in the financial crisis my estimated ROT1C already accounts to some degree the fact that Citi's business model is more risky. It can certainly be argued that maybe my model doesn't reflect the extent to which a Citigroup or JPM would be hurt by a truly calamitous unwinding of the European Union, which may be even worse than the 2008 financial crisis, but I do feel that my 10yr dataset does do a reasonable job of capturing a difficult credit cycle for the banks. Hope that's helpful and thanks for your comment.
Why Amazon's Worth North Of $200 Per Share [View article]
Super thoughtful reply Brian. Really appreciate it.
So just a couple points of clarification. If you're discounting free cash flows (and not net income), FCF adds back the stock-based comp charge that is included in net income. Thus the FCF number doesn't have a stock-based comp charge. If you've adjusted for this by discounting FCF with a stock-based comp charge included, then you're exactly right that you don't need to adjust the share count. However if you're forecasting a FCF number that does not include a charge for stock-based comp, you do need to augment the share count. I can't quite see which one you're doing from your article, but you're correct in that either approach works.
Also while you're correct that the firm will receive cash when stock options are exercised I believe most of the firm's stock-based comp is in restricted shares. When these shares vest the firm will receive no cash. So in this case, I think the cash the firm could potentially receive is much smaller given their bias to restricted stock vs option grants.
And then finally, re: including the cost of acquisitions as part of capex. You are certainly right, you can include it as a "CapEx" expense but then you need to give them credit for future growth. Or you can exclude it and not give them credit for future growth. Like you said it's one or the other, but not both. I agree with that entirely. Here's the thing though, I think their past growth benefited from acquisitions they've done in the past and since some analyst (me included) use the past as a guide to estimating the future, I think part of AMZN's growth rate is at least partially attributable to these acquisitions. So if you assume they stop acquiring, you need to assume the growth rate slows. And that it slows beyond what it was going to slow anyway given their increased size.
Hope that helps clarify my view and again thanks for the very thoughtful reply above.
Why Amazon's Worth North Of $200 Per Share [View article]
Hey Paulo, interesting point. When I look at AMZN's CapEx I include not just purchases of PP&E but cash used for acquisitions. Given that many investors are holding the stock looking for revenue growth not bottom line growth, management is essentially incentivized to purchase growth by buying other firms with steep revenue growth curves regardless of their potential to add to the bottom-line.
Another interesting thing about AMZN (and many tech companies) is that given that they pay many of their employees with stock-based comp this non-cash expense is added back when calculating free cash flow. The true effect of stock comp is to increase the share count down the road. However, since most analyst use a single share count in their DCF (the latest period), the future growth in share count from employee stock comp gets entirely lost. The way to compensate for this is to use a different share count for each future period, essentially discounting back FCF/share for each period (not several periods of FCF and then dividing by the current share count). Very few analysts do this and thus often overstate the value of firms that are heavy share comp users. Hope that makes sense.
Valuentum, interesting thorough approach. Thanks for writing.
Why LinkedIn's Stock Price Will Eventually Lose Altitude [View article]
Thanks for the article Bret. I've been less successful on my LNKD short to date, I think largely because these Growth At Any Price (GAAP) stocks can stay disconnect from valuation until there is a crack in the story. However, once there's that initial crack it's time to really add to your short. I think this data breach may be the crack that gets investors refocused on fundamentals. Time will tell.
U.S. Bancorp's Greatness Is Priced In [View article]
Hey Vireoman, thanks for your question and for reading. You're right that their business models are different in that Citi has substantially more non-US exposure than USB and that certainly can cut both ways. Having said that, Citi has always had a lot more global exposure so my return on tier 1 common should reflect the different earning potential and RISKS between these two business models. Again, my model essentially assumes that the past ten years (which include a major financial collapse) are indicative of the next ten years. Interestingly enough, because Citi faired so poorly in the last financial crisis and USB so well, my model is likely to actually overstate the safety of USB and understate the safety of Citi. This is because my expected long-term sustainable ROT1C for Citi may actually be too low if the next ten years are less disruptive and on the flip side, my long-term sustainable ROT1C for USB, could be too high if the next crisis actually manages to ensnare them.
Great question though vireoman and thanks for reading.
U.S. Bancorp's Greatness Is Priced In [View article]
I'm in the process of writing an updated BAC article which I think will explain why it's not the number 4 bank on my list. Really great catch though. I hope my updated BAC article will explain.
Wells Fargo: 3rd Best Large-Cap Bank Stock Offering Mid-Teen Returns And Limited Downside [View article]
Excellent, Excellent question. I'll be writing a new updated BAC article this week which will update my view and provide a little nuance on why I'm not recommending BAC as the number 3 bank, but why I'm still long it. Stay tuned.
U.S. Bancorp's Greatness Is Priced In [View article]
Thanks for your comment E Nuff Sed.
You're certainly right that the P/T1C ratio may not be normal and while the curves I've drawn are roughly normal, I have tried to fatten the left tail a bit to capture the asymmetric risk which exist in any regulated and levered entity like a bank. If you think the left tail should be fatter, please feel free to tweak my model, which I've shared, and see how that changes the overall return profile. I personally didn't find US Bancorp's risk profile that attractive (at least on a relative basis to the other banks), so fattening that left tail further will probably just make it look even less attractive. To your second point that I can't really know what risks are really hidden on the banks balance sheet, you are 100% correct. I don't, and likely very few people, maybe not even the CRO of these banks, truly understands all of the risks inherent in their masive balance sheets. Having said that, I make the general assumption that the over-the-cycle earnings these banks were able to generate over the last ten years, adjusted for lower leverage, is a reasonable approximation for what they can earn going forward. Basically, you can hide bad assets on your balance sheet, but you can't do it for ten years. So at some point you'll take a write-down and since I'm using GAAP earnings, not adjusted operating earnings, my analysis should capture it. So the breakdown could happen if you think the next ten years will be worse than the previous ten years. It's certainly possible that the next ten years are worse, but I'm making the assumption that they look similar to the last ten years which included one of the worst financial collapses we've seen in over 70 years. I think that's reasonably conservative, but may be not conservative enough for some. Take a look at my model and feel free to stress it for a large one-time loss and a large capital raise and see what that does to the firm's return profile. Again, I'm not a big fan of USB as I don't think there's enough of a margin of safety given the valuation, but I am recommending two of the other large cap banks Citi & JP Morgan, which have the same risk (in fact probably more risks) as USB. Yet, I think there's enough of a cushion if I've over-estimated the firm's near-term earnings for patient long-term investors. I hope that helps clarify my view and thanks for your comment. BTW, I see from your profile that you're a value investor like me. I know many value investors shun financials, so you're in great company for rejecting these types of firms outright. I however, think there is real value in some of these large-cap bank stocks despite the risks.
Own JPMorgan For The Second Best Risk/Reward Profile In The Banking Sector [View article]
Thanks for your reply E.D. Hart.
I think reasonable people can disagree whether the past 10yrs in this case (again, making some adjustments for leverage) is a good predictor of the future. I certainly agree with you that as a country we are likely to see a reasonably protracted period of global de-leveraging and that will no doubt affect the ability of these banks to grow. Having said that even if they are unable to re-invest in their business, they should still be able to return capital to investors & repurchase their stocks at a discount to book value which can continue to increase their earnings (and dividends) on a per share basis at a double-digit rate.
Also while there are some similarities btwn the Japanese period of de-leveraging over the last two decades and the US's next decade, I think one big difference is that the US injected huge amounts of capital into the banks essentially preventing them from becoming zombie banks like we had in Japan. The banking system in the US is pretty well capitalized (with a few exceptions) and likely to become more so over the next few years. Part of what I like about these investments is that I don't need to make heroic growth assumptions or predict outrageous multiples to get a pretty decent return on my money. In fact my forecasted multiples and returns on capital are well below what the banks have generated in the past. In Citi's case for example, I'm just predicting a return on tier 1 common equity of 10.25% and a price-to-tier 1 common of only 1.1x, those are far from a very aggressive assumption and well below their historical numbers. And even if you think my assumptions are too aggressive, in my sensitivity analysis, I ran a scenario where I assume the firm can only earn an 8% ROT1C and adjusted my P/T1C lower accordingly and I could still generate an IRR in the mid-teens with relatively limited downside. So even if you do believe that I've over-estimated the firm's long-term returns in a global de-leveraging cycle, I think the returns you can get on the stocks from this level, still look compelling assuming you are willing to wait.
At least in my view, the current price bakes in a pretty good margin of safety. Hope that's helpful.
Own JPMorgan For The Second Best Risk/Reward Profile In The Banking Sector [View article]
Hey Andrew, glad you enjoyed the piece. The probability distributions are purely my own estimates. Said another way, there isn't a whole lot of science behind them, but I tried to incorporate the following general premises when creating them:
1. Assumed the left tail was generally fatter than the right. This basically implies that as levered & regulated companies, there is a greater likelihood that they trade at a lower multiple to T1C than a high one. It also just makes my estimate a little more conservative.
2. Overtime, I have the weighted average P/T1C gravitate towards my calculated long-term sustainable P/T1C. In JPM's case that's 1.4x. This is much lower than what it's been historically, but it should be lower because capital requirements are going way higher.
Hope that helps. Also since these are entirely my own estimates and maybe not consistent with how other people view the probability distributions, anyone who wants can download my model here and tweak the probability distributions as they see fit. You can download my model here - http://bit.ly/LasdMV
Thanks for your question Andrew and thanks for making it all the way through a VERY long and dense piece!
U.S. Bancorp's Greatness Is Priced In [View article]
P/T1C (using Friday's close)
C 0.72x
JPM 1.04x
USB 2.48x
As you can see USB is trading at a substantial premium currently. I'm also not calling for the gap to close entirely. Again, as I believe USB's business model is lower risk and therefore deserves a premium. Below I show my targeted long-term P/T1C for each of these banks:
Long-term Sustainable Target P/T1C
C 1.1x
JPM 1.4x
USB 2.7x
However, when you look at what that implies in terms of price appreciation even if we assume no growth in Tier 1 common from earnings retention, we'd expect the following appreciation in price:
Price Appreciation from Achieving Tgt P/T1C Ratio Immediately
C +53%
JPM +35%
USB +9%
As I see it, the fact that the stock is trading at a discount to my target multiple is part of what's giving me the margin of safety.
One final point, in one respect since I'm deriving my ROT1C by looking over the last ten years and since Citi faired so poorly in the financial crisis my estimated ROT1C already accounts to some degree the fact that Citi's business model is more risky. It can certainly be argued that maybe my model doesn't reflect the extent to which a Citigroup or JPM would be hurt by a truly calamitous unwinding of the European Union, which may be even worse than the 2008 financial crisis, but I do feel that my 10yr dataset does do a reasonable job of capturing a difficult credit cycle for the banks.
Hope that's helpful and thanks for your comment.
Why Amazon's Worth North Of $200 Per Share [View article]
So just a couple points of clarification. If you're discounting free cash flows (and not net income), FCF adds back the stock-based comp charge that is included in net income. Thus the FCF number doesn't have a stock-based comp charge. If you've adjusted for this by discounting FCF with a stock-based comp charge included, then you're exactly right that you don't need to adjust the share count. However if you're forecasting a FCF number that does not include a charge for stock-based comp, you do need to augment the share count. I can't quite see which one you're doing from your article, but you're correct in that either approach works.
Also while you're correct that the firm will receive cash when stock options are exercised I believe most of the firm's stock-based comp is in restricted shares. When these shares vest the firm will receive no cash. So in this case, I think the cash the firm could potentially receive is much smaller given their bias to restricted stock vs option grants.
And then finally, re: including the cost of acquisitions as part of capex. You are certainly right, you can include it as a "CapEx" expense but then you need to give them credit for future growth. Or you can exclude it and not give them credit for future growth. Like you said it's one or the other, but not both. I agree with that entirely. Here's the thing though, I think their past growth benefited from acquisitions they've done in the past and since some analyst (me included) use the past as a guide to estimating the future, I think part of AMZN's growth rate is at least partially attributable to these acquisitions. So if you assume they stop acquiring, you need to assume the growth rate slows. And that it slows beyond what it was going to slow anyway given their increased size.
Hope that helps clarify my view and again thanks for the very thoughtful reply above.
Why Amazon's Worth North Of $200 Per Share [View article]
HP: Key In On Whitworth, Not Whitman [View article]
Why Amazon's Worth North Of $200 Per Share [View article]
Another interesting thing about AMZN (and many tech companies) is that given that they pay many of their employees with stock-based comp this non-cash expense is added back when calculating free cash flow. The true effect of stock comp is to increase the share count down the road. However, since most analyst use a single share count in their DCF (the latest period), the future growth in share count from employee stock comp gets entirely lost. The way to compensate for this is to use a different share count for each future period, essentially discounting back FCF/share for each period (not several periods of FCF and then dividing by the current share count). Very few analysts do this and thus often overstate the value of firms that are heavy share comp users. Hope that makes sense.
Valuentum, interesting thorough approach. Thanks for writing.
+ Chris
Why LinkedIn's Stock Price Will Eventually Lose Altitude [View article]
U.S. Bancorp's Greatness Is Priced In [View article]
Great question though vireoman and thanks for reading.
U.S. Bancorp's Greatness Is Priced In [View article]
U.S. Bancorp's Greatness Is Priced In [View article]
Wells Fargo: 3rd Best Large-Cap Bank Stock Offering Mid-Teen Returns And Limited Downside [View article]
Very sharp BEA!
U.S. Bancorp's Greatness Is Priced In [View article]
You're certainly right that the P/T1C ratio may not be normal and while the curves I've drawn are roughly normal, I have tried to fatten the left tail a bit to capture the asymmetric risk which exist in any regulated and levered entity like a bank. If you think the left tail should be fatter, please feel free to tweak my model, which I've shared, and see how that changes the overall return profile. I personally didn't find US Bancorp's risk profile that attractive (at least on a relative basis to the other banks), so fattening that left tail further will probably just make it look even less attractive.
To your second point that I can't really know what risks are really hidden on the banks balance sheet, you are 100% correct. I don't, and likely very few people, maybe not even the CRO of these banks, truly understands all of the risks inherent in their masive balance sheets. Having said that, I make the general assumption that the over-the-cycle earnings these banks were able to generate over the last ten years, adjusted for lower leverage, is a reasonable approximation for what they can earn going forward. Basically, you can hide bad assets on your balance sheet, but you can't do it for ten years. So at some point you'll take a write-down and since I'm using GAAP earnings, not adjusted operating earnings, my analysis should capture it. So the breakdown could happen if you think the next ten years will be worse than the previous ten years. It's certainly possible that the next ten years are worse, but I'm making the assumption that they look similar to the last ten years which included one of the worst financial collapses we've seen in over 70 years. I think that's reasonably conservative, but may be not conservative enough for some. Take a look at my model and feel free to stress it for a large one-time loss and a large capital raise and see what that does to the firm's return profile. Again, I'm not a big fan of USB as I don't think there's enough of a margin of safety given the valuation, but I am recommending two of the other large cap banks Citi & JP Morgan, which have the same risk (in fact probably more risks) as USB. Yet, I think there's enough of a cushion if I've over-estimated the firm's near-term earnings for patient long-term investors.
I hope that helps clarify my view and thanks for your comment. BTW, I see from your profile that you're a value investor like me. I know many value investors shun financials, so you're in great company for rejecting these types of firms outright. I however, think there is real value in some of these large-cap bank stocks despite the risks.
U.S. Bancorp's Greatness Is Priced In [View article]
Own JPMorgan For The Second Best Risk/Reward Profile In The Banking Sector [View article]
I think reasonable people can disagree whether the past 10yrs in this case (again, making some adjustments for leverage) is a good predictor of the future. I certainly agree with you that as a country we are likely to see a reasonably protracted period of global de-leveraging and that will no doubt affect the ability of these banks to grow. Having said that even if they are unable to re-invest in their business, they should still be able to return capital to investors & repurchase their stocks at a discount to book value which can continue to increase their earnings (and dividends) on a per share basis at a double-digit rate.
Also while there are some similarities btwn the Japanese period of de-leveraging over the last two decades and the US's next decade, I think one big difference is that the US injected huge amounts of capital into the banks essentially preventing them from becoming zombie banks like we had in Japan. The banking system in the US is pretty well capitalized (with a few exceptions) and likely to become more so over the next few years. Part of what I like about these investments is that I don't need to make heroic growth assumptions or predict outrageous multiples to get a pretty decent return on my money. In fact my forecasted multiples and returns on capital are well below what the banks have generated in the past. In Citi's case for example, I'm just predicting a return on tier 1 common equity of 10.25% and a price-to-tier 1 common of only 1.1x, those are far from a very aggressive assumption and well below their historical numbers. And even if you think my assumptions are too aggressive, in my sensitivity analysis, I ran a scenario where I assume the firm can only earn an 8% ROT1C and adjusted my P/T1C lower accordingly and I could still generate an IRR in the mid-teens with relatively limited downside. So even if you do believe that I've over-estimated the firm's long-term returns in a global de-leveraging cycle, I think the returns you can get on the stocks from this level, still look compelling assuming you are willing to wait.
At least in my view, the current price bakes in a pretty good margin of safety. Hope that's helpful.
Revisiting The Intrinsic Value Of Berkshire Hathaway [View article]
Own JPMorgan For The Second Best Risk/Reward Profile In The Banking Sector [View article]
1. Assumed the left tail was generally fatter than the right. This basically implies that as levered & regulated companies, there is a greater likelihood that they trade at a lower multiple to T1C than a high one. It also just makes my estimate a little more conservative.
2. Overtime, I have the weighted average P/T1C gravitate towards my calculated long-term sustainable P/T1C. In JPM's case that's 1.4x. This is much lower than what it's been historically, but it should be lower because capital requirements are going way higher.
Hope that helps. Also since these are entirely my own estimates and maybe not consistent with how other people view the probability distributions, anyone who wants can download my model here and tweak the probability distributions as they see fit. You can download my model here - http://bit.ly/LasdMV
Thanks for your question Andrew and thanks for making it all the way through a VERY long and dense piece!