Blind Faith Price Targets

[jhm]

Hmm, I like the simplicity of market cap for implied discounts/returns. It does a nice job of showing the stability of a distribution with mean reversion, and this is technically what would make the BFPT methodology work.

Still I think there is one teensie problem. The return on a share of stock is not the same as the return on market cap (unless you factor in capital raised or distributed as the float increases or decreases). In other word the gap between the return on a share and the return on market cap is dilution. Thus, while Tesla is prone to raise capital the market will put a higher discount on shares than it would on market cap.

From a religious point of view, our ecumenism was tested by dilution factors. Then the Classic Musk Blind Faithful threatened to reject Reformed Blind Faith and return to fundamentals. Then the heavens open with word from Tesla on a new covenant with Elon, whence a movement to Blind Faith Renewal opened the prospect of a restored union of the Blind. But alas we return to the thorny issue of dilution. We're seeing millennia of religious history play out in a matter of days. We must be getting close to the Singularity!

 

[jhm]

If the goal is to predict where the stock is going to be in the future, yes, you need to worry about dilution.

However, if you are simply being reactive, as in trying to answer the question of where is the stock right now against history and expectations, market cap is the right approach. Looking to lighten up or write covered calls? Do it when the discount is on the lower end. Looking to add more or write puts? do it when the discount is on the upper end.

Yeah, we're not really trying to predict prices. But the point more is to track progress to a long term target and the get a sense of sentiment a long the way. Did you see the back test a few days back. Buying at the low end of sentiment (high implied discout) seems to have led to superior annual returns.

In any case, we are looking at dilution rates that are so thin, it really does not make much differnce in near term Pts. The sensitivity to choice of LTPT is low so long as the target is sufficiently far out in the future. And that is one reason why I like to keep it ten years out. The key here is that we are calibrating on recent prices. So long as the LTPT is far enough out it is not so influent relative to recent market prices. As we get close to a LTPT then implied discounts can diverge to a rally high value or low value, even negative value. In fact they will only converge if the market price exactly hits the LTPT. So again that is why I want to keep that target well out in the future. Do that an most reasonable choices of LTPT will give you a reactive tool. It is good to look at a plot of implied discounts to be sure that it is not trending too strongly in one direction or the other.

 

[ajax]

For the dilution conversation:

Chamath Palihapitiya, who is a silicon valley VC and with reach to Musk, said that- management told him that they will need 12 to 15Bil to execute the plan. Then he added that he thinks it's more like 20Bil (knowing that Musk is always optimistic with his projections).

This is in addition to the 1% dilution for Musk pay and maybe another 1% for the rest of Tesla employees.

Where did Chamath say that out of curiosity? My google-fu skills have failed me.

Was the $12-15 billion needed on top of their expected cash flow?

 

[bdy0627]

It is interesting to think about how the chart above shows roughly 4 clusters of observations.

Top cluster, sentiment >85%: here the market is very bullish pushing the limits of how fast the stock can grow. So there may be an opportunity for a trend play. This might become more clear if we looked at a shorter term return. At any rate within this cluster there is a downward slope suggesting that getting in early, near 90%, is better than late, above 95%. If timed perfectly a 30% return is possible, but timed badly risks a -20% return.

High cluster, sentiment 55% to 80%: Here we have a fairly stagnant market. The downside seems a little more heavily weighted than the upside. This is likely to be a mix of uptrend and down trends. So there may be ways to trade this, but generally the stock may be priced just a bit on the optimistic side. So there is always the risk that sentiment could deteriorate.

Low cluster, sentiment 35% to 50%: this bearish market seems to be transitional. The upside appears to be more heavily weighted than downside. Well timed, over 30% returns are posible, but poorly timed still risks a -15% return. Accumulation here may be a reasonable multi-year investment.

Bottom cluster, sentiment < 30%: The stock is truly out of favor and trading at a discount in this market. The price seems to be bottomed out. The upside is 40% to 85% return, practically no downside. This is back the truck up time, because the market would seem to be irrationally bearish on Tesla. In the other clusters, one needs to be a savvy trader to avoid a sharp loss, but in the bottom market one simply needs to buy and hold and have enough heart to battle enormous negativity. This is the blind faith zone.

This is really interesting and of course not surprising. Is your sentiment rating based upon the BFPT percentile or discount at each time interval?

 

[bdy0627]

It is interesting to think about how the chart above shows roughly 4 clusters of observations.

Top cluster, sentiment >85%: here the market is very bullish pushing the limits of how fast the stock can grow. So there may be an opportunity for a trend play. This might become more clear if we looked at a shorter term return. At any rate within this cluster there is a downward slope suggesting that getting in early, near 90%, is better than late, above 95%. If timed perfectly a 30% return is possible, but timed badly risks a -20% return.

High cluster, sentiment 55% to 80%: Here we have a fairly stagnant market. The downside seems a little more heavily weighted than the upside. This is likely to be a mix of uptrend and down trends. So there may be ways to trade this, but generally the stock may be priced just a bit on the optimistic side. So there is always the risk that sentiment could deteriorate.

Low cluster, sentiment 35% to 50%: this bearish market seems to be transitional. The upside appears to be more heavily weighted than downside. Well timed, over 30% returns are posible, but poorly timed still risks a -15% return. Accumulation here may be a reasonable multi-year investment.

Bottom cluster, sentiment < 30%: The stock is truly out of favor and trading at a discount in this market. The price seems to be bottomed out. The upside is 40% to 85% return, practically no downside. This is back the truck up time, because the market would seem to be irrationally bearish on Tesla. In the other clusters, one needs to be a savvy trader to avoid a sharp loss, but in the bottom market one simply needs to buy and hold and have enough heart to battle enormous negativity. This is the blind faith zone.

If I am understanding all of this correctly, assuming the 4% dilution model, we are currently around 27% percentile/32% discount. This places us in the bottom sentiment cluster, correct?

 

[jhm]

This is really interesting and of course not surprising. Is your sentiment rating based upon the BFPT percentile or discount at each time interval?

It is the percentile of the discount, not the discount directly. I've looked at the latter, but it does not show as strong an indicator. So something about contextuallizing discount to recent history seems to be helpful for finding the bottom.

 

[jhm]

If I am understanding all of this correctly, assuming the 4% dilution model, we are currently around 27% percentile/32% discount. This places us in the bottom sentiment cluster, correct?

Yes.

 

[jhm]

The current 32% percentile comes in at $357. So we could easily slip out of the lower cluster in the next few days.

 

[bdy0627]

The current 32% percentile comes in at $357. So we could easily slip out of the lower cluster in the next few days.

Thanks. It doesn't "feel" like we are in that lowest sentiment group at this point, but I guess that's because we only barely are.

 

[jhm]

Thanks. It doesn't "feel" like we are in that lowest sentiment group at this point, but I guess that's because we only barely are.

I agree. On Dec 27, 2017, we were down to the 12% percentile at a price of $312. That was our most recent bottom. Also on Dec 5, we were at the 11% at $304. So it pays to watch out for opportunities like that.

Who knows? Maybe we get another opportunity to buy at $312, or we could be headed to $400 with no going back.

The opportunities to buy at below the 12% percentile are actually quite rare. That is the lower values are usually in the past where you can't trade them. So to be in the 12% percentile in the present moment is a rare thing.

 

[adiggs]

I know you've posted the spreadsheet and info before @jhm, but would you post again the spreadsheet and details to enable a person (such as myself) to update the BFPT on my own and at my whim?

A challenge with seeing a really low discount in the moment is it's got to be calculated each day (or at least frequently), and there's no real need to have you do it each time somebody gets inspired to ask you to do so

Thanks!

 

[bdy0627]

I think it would be great to post an alert whenever we drop into the most bearish sentiment group. As we move forward, and the prices go up, it's easy to lose sight of where that is. That's a really valuable piece of information. Of course, it only works until it doesn't, but I'm ok with that.

 

[jhm]

I know you've posted the spreadsheet and info before @jhm, but would you post again the spreadsheet and details to enable a person (such as myself) to update the BFPT on my own and at my whim?

A challenge with seeing a really low discount in the moment is it's got to be calculated each day (or at least frequently), and there's no real need to have you do it each time somebody gets inspired to ask you to do so

Thanks!

Actually, I have not shared my spreadsheet, but would recommend that others build their own. Or if someone wants to program an app or web tool, that would be cool too. In either case, the virtue of building your own is that you become much more aware of how it works. Some people have done this.

The price targets do not change much from day to day. So updating daily is not really needed. For example, the current 10% and 30% percentile price targets are $310 and $349 and in 30 days they'll be $317 and $357. (You can also advance these using the discount rates.) So buying under $350, but not waiting for it to go down to $310 is a fairly good rule for the next month for accumulating in the bottom cluster.

Here's a little mini-update using the $5455 LTPT. Note I've changed the percentile points to report.

Percentile    Discount    2018-01-25    2018-02-24    2018-12-31    2019-02-24    2019-12-31    2020-12-31    2022-12-31    2025-12-31
26.8%    32.02%     $345      $353      $447      $466      $590      $780      $1,359      $3,130
0%    35.93%     $258      $265      $344      $360      $467      $635      $1,174      $2,952
10%    33.46%     $310      $317      $406      $424      $541      $723      $1,287      $3,063
30%    31.86%     $349      $357      $452      $471      $596      $787      $1,368      $3,137
50%    30.08%     $400      $409      $511      $531      $664      $865      $1,464      $3,224
70%    28.24%     $461      $470      $581      $603      $745      $956      $1,572      $3,317
90%    26.56%     $525      $535      $654      $677      $827      $1,048      $1,679      $3,405
100%    24.75%     $606      $617      $745      $770      $929      $1,159      $1,804      $3,505

 

[jhm]

I've been playing around with my backtesting on sentiment. I was originally using a 4 year look back to get the percentiles. When I change that to 2 or 3 years instead, I different threshold between the low cluster and above. So there is a fair amount of sensitivity to this assumption on how much history to calibrate to.

This sensitivity is a turnoff for me. So I am taking a fresh look at simply using implied discount instead. The upshot here is that there is no need to make any assumption about how far to look back. So here is how the backtest on discounts looks.

Notice that the trend is fairly strong and has a little curvature to it, hence, the quadratic fit. We don't have as much clustering as we saw with sentiment. So it is less clear where exactly to draw a line between where to buy and where to hold or even sell. It seems that a reasonable buy limit is somewhere between 31% and 32%. So maybe 31% is a good threshold, where mean return has been in excess of 20%. On the sell side a threshold in 28% to 30% could make sense. Below 28% discount the mean return has been negative.

So my rough trading rule would be like accumulate when discount is over 31% and lighten position when discount is below 29%. Of course, there are a zillion other factors to consider when actually trading.

Warning, these thresholds are specific to out $5455 LTPT at end of 2027. Any change of LTPT would require a different backtest to obtain suitable thresholds.

 

[jhm]

In case you were wondering...

Action    Discount 2018-01-25 2018-02-24 2018-12-31 2019-02-24 2019-12-31
 Strong Buy     32%     $346      $354      $448      $467      $591
        Buy     31%     $373      $381      $479      $499      $628
       Hold     30%     $402      $411      $514      $534      $668
       Sell     29%     $434      $444      $551      $572      $710
Strong Sell     28%     $469      $479      $591      $613      $756

 

[jhm]

Hmm, maybe I want to change my extreme cut points a bit. Look at this tabulation of the 771 day backtest.

Discount  Count  AverageAnnualReturn  StdDevAnnualReturn
24% to 27%    155    -2.9%    13%
27% to 29%    225    -2.2%    16%
29% to 31%    158     4.3%    18%
31% to 33%    116     44.1%    16%
33% to 36%    117     66.5%    13%
Grand Total   771     16.4%    31%

Corresponding to these cuts...

Action Discount 2018-01-25 2018-02-24 2018-12-31 2019-02-24 2019-12-31 2020-12-31
 Strong Buy    33%     $321      $328      $418      $437      $556      $740
        Buy    31%     $373      $381      $479      $499      $628      $823
       Hold    30%     $402      $411      $514      $534      $668      $869
       Sell    29%     $434      $444      $551      $572      $710      $917
Strong Sell    27%     $507      $517      $634      $657      $805     $1023

So give me some feedback on these cut points. How well do they align with your personal thresholds?

Personally, I would only accumulate above 32% discount / below $346, but would probably hold out for 32% / $321. On the sell side, I am very reluctant to sell at nearly any price. Even if the stock loose 10% over the next 12 months, I would be inclined to ride it out. But if the stock really were rise to near $500 within the next month or two, I would want to lighten up. If nothing more than to have some cash for when the price pulls back. So somewhere in the 27% to 28% range of discount, I'd want to lighten up.

 

[SBenson]

Where did Chamath say that out of curiosity? My google-fu skills have failed me.

Was the $12-15 billion needed on top of their expected cash flow?

Yes, on top of expected cash flow. He said it at Sohn conference but I am unable to find it on internet. Pretty sure I saw it on a pro-tool. Will send you a screenshot in PM that is relevant.

 

[bdy0627]

Hmm, maybe I want to change my extreme cut points a bit. Look at this tabulation of the 771 day backtest.

Discount  Count  AverageAnnualReturn  StdDevAnnualReturn
24% to 27%    155    -2.9%    13%
27% to 29%    225    -2.2%    16%
29% to 31%    158     4.3%    18%
31% to 33%    116     44.1%    16%
33% to 36%    117     66.5%    13%
Grand Total   771     16.4%    31%

Corresponding to these cuts...

Action Discount 2018-01-25 2018-02-24 2018-12-31 2019-02-24 2019-12-31 2020-12-31
 Strong Buy    33%     $321      $328      $418      $437      $556      $740
        Buy    31%     $373      $381      $479      $499      $628      $823
       Hold    30%     $402      $411      $514      $534      $668      $869
       Sell    29%     $434      $444      $551      $572      $710      $917
Strong Sell    27%     $507      $517      $634      $657      $805     $1023

So give me some feedback on these cut points. How well do they align with your personal thresholds?

Personally, I would only accumulate above 32% discount / below $346, but would probably hold out for 32% / $321. On the sell side, I am very reluctant to sell at nearly any price. Even if the stock loose 10% over the next 12 months, I would be inclined to ride it out. But if the stock really were rise to near $500 within the next month or two, I would want to lighten up. If nothing more than to have some cash for when the price pulls back. So somewhere in the 27% to 28% range of discount, I'd want to lighten up.

These line up pretty closely to where I like adding. I am generally adding calls, so I'm fairly cautious about having it be down before adding. I hold on to the bulk of my investment regardless of where it is, though I agree with lightening up if it were to get up near $500 in the next few months. For me currently, I'm adding below $340 with the 200 MA around $335.

 

[bdy0627]

According to the discounts, I would definitely not add at 31%. I would definitely add at 33% and probably at 32.5%. There is a lot of room between 31 and 33, so fractions matter here.

 

[EinSV]

Hmm, maybe I want to change my extreme cut points a bit. Look at this tabulation of the 771 day backtest.

Discount  Count  AverageAnnualReturn  StdDevAnnualReturn
24% to 27%    155    -2.9%    13%
27% to 29%    225    -2.2%    16%
29% to 31%    158     4.3%    18%
31% to 33%    116     44.1%    16%
33% to 36%    117     66.5%    13%
Grand Total   771     16.4%    31%

Corresponding to these cuts...

Action Discount 2018-01-25 2018-02-24 2018-12-31 2019-02-24 2019-12-31 2020-12-31
 Strong Buy    33%     $321      $328      $418      $437      $556      $740
        Buy    31%     $373      $381      $479      $499      $628      $823
       Hold    30%     $402      $411      $514      $534      $668      $869
       Sell    29%     $434      $444      $551      $572      $710      $917
Strong Sell    27%     $507      $517      $634      $657      $805     $1023

So give me some feedback on these cut points. How well do they align with your personal thresholds?

Personally, I would only accumulate above 32% discount / below $346, but would probably hold out for 32% / $321. On the sell side, I am very reluctant to sell at nearly any price. Even if the stock loose 10% over the next 12 months, I would be inclined to ride it out. But if the stock really were rise to near $500 within the next month or two, I would want to lighten up. If nothing more than to have some cash for when the price pulls back. So somewhere in the 27% to 28% range of discount, I'd want to lighten up.

The Strong Buy and Buy cut points make sense to me and the data you have compiled suggest good results so far with that approach.

Personally I would not use the model as guidance for "sell" levels and instead would think of those levels as something like a "cautionary" signal for short-term moves. I have explained my thinking before (posts 164, 166, 168, 223, 225) but long story short even a 27% discount rate, which is a "Strong Sell," is extremely high and would be difficult to match with other investments.

I continue to believe that if Tesla executes on the Model 3 ramp (including gross margin targets) and barring macro hiccups or other surprises we are likely to see an inflection point between now and 2020 (as we had in 2013) where the SP pops and the implied discount rate to the LTPT drops from ~30% to a more "reasonable" number -- perhaps something closer to 20%. I'd hate to sell and miss out on the fun.