Category: Tools

Posted on by — Leave a comment

#6: TOOLS – Want to know where the S&P500 will most likely close tomorrow, based on the recent price action?

🤖 This code retrieves a series of price data (for an index or stock, you decide) going back X number of periods, with Y timeframe resolution.

It looks at the Z most recent data points, cycles through the whole dataset to find windows that are Z periods long and have at least A correlation with the most recent data.

It picks the B number of periods with the highest correlation, then looks at how price developed in the following C number of periods, then applies the averge of that development to the most recent datapoint (i.e. yesterday’s close) to get a price prediction for C periods into the future.

The code prompts you to set values for index, X, Y, Z, A,B and C yourself when you run it.

🤖 This python code runs on Google Colab (google’s free online editor, kind of a google docs for code) so you can run it directly there, or just copy paste into your code editor of choice.

So no coding required.

🤖 The code retrieves prices from Yahoo Finance via an API. It’s free and doesn’t require any login.

Yahoo Finance has 20+ years of free daily price data for most indexes (over 40 years for the S&P500), but not very much for lower time frames like 60m, 15m or 1m.

But you can easily adjust the code (or ask ChatGPT to help you) retrieve data from better data suppliers like AlphaVantage.

🤖 So for example, based on the past 80 days price action, the code looks for the 10 most similar periods in the past 10,000 periods (about 40 years) that shows >0.75 correlation.

The average performance following those periods suggests that the S&P will be down 1.7% in the coming 10 days.

❗️Note that this is for information purposes only and not to be relied upon for investment decisions.

Every moment in the market is unique.

❗️❗️ Please note that I can not perform any support on running or modifying this code. Just ask ChatGPT.

RUN OR MAKE A COPY OF THE CODE ON LINK BELOW

You can run the code on this Google Colab link, but to edit you have to make a copy of it or copy paste the code into your editor of choice.

THIS RESOURCE REQUIRES A FREE SUBSCRIPTION. IF YOU WANT TO VIEW THIS RESOURCE PLEASE SIGN UP FOR A FREE SUBSCRIPTION HERE

Posted on by — Leave a comment

#5: TOOLS – What is your favorite biotech stock worth when it’s time for a Big Pharma license deal?

💊 The template can easily be adapted to raw material exploration projects. Or any risk-adjusted project with heavy investments upfront and revenues far ahead in the future.
 
Also, prints nicely to one horizontal A4.
 
💊 Most biotech R&D companies have no ambition to bring their projects all the way to the market.
 
Instead they target to get through Phase I and II, to proof-of-concept and then get a license deal with a Big pharma company.
 
💊 With first revenues sometime 7-10 year into the future, it’s very difficult to value such assets.
 
How do licensor and licensee come up with these sometimes astronomical sums in some license deal?
  
💊 A biotech license deal model must adjust for the varying stages of likelihood of success.
 
At the beginning of clinical development, there might be a 63% to get through Phase I and 25% chance for a successful Phase II. That means that the cumulative probability to get from beginning of Phase I to a successful Phase II is 63% x 25% = 16%. 
 
On the flip side, after a successful Phase I there is a 100% probability that you will take the cost for Phase II but only a 25% probability that you will take the cost for Phase III.
 
The model needs to take these cumulative probabilities into account.  
 
Luckily, there are good statistics on recent success rates for various types of clinical development so you can get a pretty good idea. (link in the excel)
 
💊 Some key challenges in modeling include clinical development costs, time-frame and sales forecast profile.
 
Some features of this model include
 
– Top down sales forecast based on patients in the 7MM (seven major medical markets) with forecast built around estimated peak market share. But you can of course adjust to any other metric.
 
– Calculation for different discount rates for licensee and licensor
 
– Estimates for R&D expenses, COGS and SG&A (normally 15% and 20% of sales respectively, are good estimates for this)
 
– Upfront payment and risk adjusted milestone payments
 
– Tiered Royalty rates as a function of sales volume
 
This produces one NPV for the licensor and one NPV the licensee. 
 
💊 To gauge the size of the upfront, the milestones and the royalty in relation to the full value, you have to consider the risk profile in the project and structure in similar deals.
 
Some hot niches get very high sums upfront, while other deals are more focused on compensating the licensor in milestone payments and through royalties.

DOWNLOAD THE EXCEL FILE HERE  

To edit the file you will have to download it to your computer as an .xls-file. Alternatively you can make a copy and edit in Google Sheets but that will mess up some of the formatting.

https://docs.google.com/spreadsheets/d/1wepCKxF2eB-D6wHhqQmGqSa6DK7STOYA/edit?usp=sharing&ouid=110084224644770502017&rtpof=true&sd=true