Updated: Jan 13, 2020
Nvidia (Symbol: NVDA) is a stock which will get you really rich if you were to invest in it in 2009 and hold it all the way till 2019/20. In fact, its growth is so phenomenal that there is actually an article showing you how much your $1000 investment in NVDA would have been in 2019 had you invested it in 2009 and hold it. In comparison, you could have made almost 5 times the return had you have invested in S&P500 in the same time horizon. This is truly amazing considering the S&P500 could have already made you 2.5 times richer if you buy and hold since 2009 till 2019.
Personally, I had great admirations for Nvidia's management as they are one of the few companies which first recognises the need to transform themselves from a hardware company to an AI company. Years before people recognise the importance of AI, Nvidia had already started to slowly build its moat around its hardware to enable developers to build deep learning models from it's CUDA library. Such foresight had certainly serve them well as they are now one of the top AI companies. This is truly remarkable if you consider that they are traditionally a hardware company. Many hardware companies did not recognise the AI trend years ago and are now struggling to catch up with the AI wave.
Now, let's take a look at NVDA's performance over the past year.
Similar to most other tech stocks, NVDA has been on an uptrend in the last few months for the past year. However, it has not seen a new high since it reached ~280USD in Sep 2018.
How does it compare against the major US indices then?
NVDA has been moving quite similarly as compared to the major US indices. Here are the r coefficients for the trend lines.
NVDA performs slightly worse than the major US indices. However, I would say that the performance is not too different.
Now, one important factor in trading stocks is also to take note of the various technical indicators. One of the commonly used indicators has been the Simple Moving Average (SMA). Using Simple Moving Average, we are able to better tell how the price trend has been like. For eg, if SMA50 (50-day Simple Moving Average) rises above SMA200 (200-day Simple Moving Average), a golden sign arises which represents a bullish signal. Similarly, if SMA50 drops below SMA200, a death cross arises which represents a bearish sign. The more such crossovers happen, the more volatile the prices are deem to be.
I personally prefer to use SMA20 and SMA50 as there is less "lag" to the actual price chart since it is shorter term in nature, and would like to see how does the crossover chart look like in comparison to the price chart.
You can see that there are five instances where the SMA20 crosses the SMA50. The first occurs in Feb 2019 where the SMA20 rises above SMA50 (bullish sign). One occurs some time in May 2019 where the SMA20 drops below SMA50 (bearish sign). The third occurs some time in Jul 2019 where the SMA20 rises above SMA50. The fourth occurs some time near Sep 2019 where the SMA20 drops below SMA50, with a sharp rise of SMA20 above SMA50 again in Sep 2019.
What happens if you buy NVDA only on 1 Feb 2019 when a bullish sign occurs (SMA20 rises above SMA50), sell it when SMA 20 drops below SMA50 and only reinvest again when the SMA20 rises above SMA50 again (rise and repeat)? Would you have made a bigger profit? Let's take a look.
If you were to buy NVDA on 7 Jan 2019 and hold it till now with an initial capital of 1000USD, your 1000USD will have turned into 1646USD.
If you were to adopt the method as described above to rely on these bullish/bearish signal, your 1000USD will have turned into 1469USD instead! In this case, buy and hold is a better strategy as compared to relying on technical indicators.
However, we have only used SMA20 and SMA50 in this case with a very short time horizon of 1 year. Results might potentially be different if you use other SMA values or extend the time horizon.
Sentiment Analysis is another area where data science has been applied to finance to better understand if people sentiments has any effect on stock prices. In this case, I will attempt to try to assign a quantitative value to people sentiments on Nvidia to see if there is a direct correlation between people sentiments and NVDA stock price.
In this case, I extracted NVDA sentiment data from Sentdex which pulls data from a variety of sources such as Reuters, Yahoo Finance, Bloomberg, Forbes etc and assigns a value to the general sentiment on a particular topic.
Here is the scale, ranging from a value of -3 (strongly negative) to 6 (strongly positive)
6 - Strongest positive sentiment
5 - Extremely strong, positive, sentiment
4 - Very strong, positive, sentiment
3 - Strong, positive sentiment
2 - Substantially positive sentiment
1 - Barely positive sentiment
0 - Neutral sentiment
-1 - Sentiment trending into negatives
-2 - Weak negative sentiment
-3 - Strongest negative sentiment.
To ensure higher accuracy in the data value, I use the Simple Moving Average of the sentiment value across a period of 5 days instead.
Here is how it looks like (as an example)
I extracted the sentiment value over a duration of 1 year, and cross examine it against NVDA stock price in a scatter plot. Here is how it looks like.
The r coefficient for the trend line is 0.707, which suggests there is a strong correlation between people sentiments and NVDA share price. As such, NVDA is a stock which you can factor in people sentiments when analysing a good entry price for the stock.
With AI becoming increasingly important and prevalent in all industries in the future, there is no doubt company like NVIDIA has a bright future ahead. However, its current PE ratio of 59.6X is way higher than the US semiconductor industry average of 33.6X. Its current PB ratio of 12.9X is also way higher than the US semiconductor industry average of 3.2X, suggesting NVIDIA might be way overvalued based on its current share price. Is Nvidia's advancement in the AI space worth this extra premium which you are paying for now?
You be the judge!
Psst.. If you like what you read, please scroll down and subscribe for regular updates!