Blog Update

Hi Everyone,

A quick update on the Blog. I recently (back in June) accepted a new position doing equity research at an asset management firm in Houston. To maintain good standing with compliance, I will no longer be posting investment ideas here.

However, my new team and I are planning on writing about some of our ideas going forward over at https://www.globescancapital.com/blog. Feel free to check it out in a few months and see what we are coming up with. As always, thanks for reading.

Best,
David Shahrestani

P.S. I passed CFA Level III!

Advertisements

Analysis: ALJ Regional Holdings, Inc. (NASDAQ: ALJJ)

ALJ Regional Holdings, Inc (NASDAQ: ALJJ) is a private-equity-style holding company, with a solid history of operations, that currently trades at a 15% free cash flow yield. The stock has suffered a 50% drawdown over the past year, mainly due to a stream of bad news, including ugly quarterly results, fears of a debt covenant breach, NOL’s taking a hit from the tax bill, and a sexual harassment lawsuit filed against the firm’s CEO and key capital allocator, Jess Ravich (in his role at TCW). The main question is does this negative sentiment match reality?

This document contains an independent analysis of the company. Information on my investment philosophy can be found here. On a side note, I haven’t written much lately as I’ve been busy studying for the CFA Level III exam and interviewing for jobs.

Continue reading “Analysis: ALJ Regional Holdings, Inc. (NASDAQ: ALJJ)”

Programming #15: NLP Classification

Previously, we explored how a Recurrent Neural Network could be used to translate French text into English text and how a Convolutional Neural Network could be used to predict a dog’s breed based on a picture. Today, we’ll be playing around with combining the two in order to solve a difficult natural language processing problem:

Given a user comment from the internet, classify whether the comment is:

  • toxic
  • severe_toxic
  • obscene
  • threat
  • insult
  • identity_hate

As always, the full code for this project can be found on my GitHub.

Continue reading “Programming #15: NLP Classification”

Investment Theory #19: Berkshire Hathaway’s 1980 Letter

This post continues our series on Berkshire Hathaway’s annual letters. Original letters can be found here.

Links to past years: 1977, 1978, 1979

Links to Buffett’s partnership years: 1957, 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968

Continue reading “Investment Theory #19: Berkshire Hathaway’s 1980 Letter”

Analysis: Q4-2017 Macro Update

Global economic expansion, falling unemployment levels, low interest rates, and low volatility have coalesced, creating a Goldilocks environment for asset valuations. This party continued throughout the fourth quarter and perhaps it will continue far into the future, as markets appear to be betting on; trees will grow to the sky, and everyone will get rich. But perhaps not.

As always, I have no idea what the near-term future will bring, which naturally leads me towards skepticism whenever expectations are at or near extremes (things are rarely as bad as they seem, and things are rarely as good as they seem). In the face of uncertainty, the best thing we can do is study up on economic history, trying to get a better sense of how today’s world fits in a larger picture. There isn’t much money to be made in trying to time the market, but there is plenty of money to be made in understanding the world a little better than the crowd.

Continue reading “Analysis: Q4-2017 Macro Update”

Analysis: Pendrell Corporation (OTCMKTS: PCOA)

Pendrell Corporation (OTCMKTS: PCOA) is an illiquid, closely held microcap which is currently trading below its net cash position. This document contains an independent analysis of the company. Information on my investment philosophy can be found here.

Disclaimer: The author is currently long PCOA. All of the views expressed in this document are solely those of the author and do not reflect the view of any other person or company. I am not receiving compensation for it. I have no business relationship with the company whose stock is mentioned in this document. All investments involve the risk of permanent capital loss. I encourage everyone to do their own due diligence and to reach their own conclusions. Under no circumstances should this document be considered an offer to buy or sell any securities mentioned within.

Continue reading “Analysis: Pendrell Corporation (OTCMKTS: PCOA)”

Year in Review: 2017

As each year ends, I try to pause and reflect on some of the year’s themes and how poorly I predicted them. The goal of this exercise is to relearn some of the lessons we relearn every year: Mainly that nobody, including myself, has any idea how macroeconomic and sociopolitical events will play out, nor how they will affect the markets. The best we can do is predict that past inertia will continue into the future – e.g. the stock market will return about 7.0%. We know this is an illusion since, within any chaotic, complex system, inertia is ephemeral at best, but it’s a nice illusion.

Continue reading “Year in Review: 2017”

Investment Theory #18: Berkshire Hathaway’s 1979 Letter

This post continues our series on Berkshire Hathaway’s performance from 1977 to today. Shareholder letters can be found here.

Links to previous posts: 1977, 1978

Links to Buffett’s partnership years: 1957, 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968

Continue reading “Investment Theory #18: Berkshire Hathaway’s 1979 Letter”

Programming #13: Recurrent Neural Networks

Last time, we explored how a Convolutional Neural Network could be trained to recognize and classify patterns in an image. With a slight modification, a CNN could also be trained to generate new images. But what if we were given a series of frames in an animation and wanted our CNN to predict the next frame? We could feed it a bunch of two frame pairs and see if it could learn that after frame ‘a’ usually came frame ‘b’ but this wouldn’t work that great.

What we really need is a neural network that is able to learn from longer sequences of data. For example, if all the previous frames show a ball flying in an arc, the neural network might be able to lean how quickly the ball is moving in each subsequent time period and make a prediction on the next frame based off that. This is where Recurrent Neural Networks (RNN) come in.

Today, we’ll be conceptualizing and exploring RNN’s by building a deep neural network that functions as part of an end-to-end machine translation pipeline. Our completed pipeline will accept English text as input and return the French translation as output. You can follow along with the code here.

Continue reading “Programming #13: Recurrent Neural Networks”