Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Sandalwood AdvisorsGenerating Alpha from big data analytics
Sandalwood Advisors 2
Sandalwood Advisors 3
Sandalwood Advisors 4
Sandalwood Advisors
• From news, clearly there is a big trend of both hedge funds, and asset managers looking for opportunities in alternative data
Quick intro: Who is Sandalwood Advisors?
Sandalwood Advisors is Asia’s largest alternative data firm by data assets. You can gain access to 7 unique sources of data via our team.
Established in 2015
5
Sandalwood Advisors
What is Alternative Data? How is it different to Traditional Data?
6
Alternative Data – any data that is not available on traditional data sources like Bloomberg, Reuters, Factset.
Difference with traditional dataset
1) Frequency – Quarterly, semi annual report vs monthly data
2) Granularity –Topline sales number vs granular analysis on consumer base
3) Reality – Company always paint themselves in best possible light
4) Scarcity – Traditional data is priced in. Very few people have access to alternative data
Sandalwood Advisors 7
Sandalwood Advisors
Top 3 reasons why investor use Alternative data.
1. Generating non-consensus views
2. Confirming Investment Thesis - getting higher conviction
3. Keeping management of the company honest
8
Sandalwood Advisors
Kering Group Gucci – Catching break-out trends early
9
Sandalwood Advisors
What are the 5 most common misconceptions of alternative data?
1. I need data scientist to use alternative dataThere are many structured datasets which investment professional can readily use
2. Alternative data is for quants or traders only, not for long term investor.A lot of alternative dataset can identify long term trends, or signals of structure shifts ahead of traditional dataset
3. PMs or analysts over-relying on a single source of data points.Many pms/analysts think one dataset is powerful enough to make investment decision, reality is you need multiple points of data to generate thesis
4. If everyone buys data, then it won’t be as powerful.There are many interpretation of a single data point
5. Dataset must be correlated with company earnings in order for it to be useful.
There are many dataset which are not correlated to company quarterly earnings, but can generate powerful investment signals
10
Sandalwood Advisors
Anta – how come there are no correlation?
11
Sandalwood Advisors
Again, data is not correlated with company earnings
12
Sandalwood Advisors 13
Sandalwood Advisors
Anta – You can use sector correlations
14
Sandalwood Advisors
Anta saw the highest growth vs other major brands in the offline channel. Nike sales also recovered. Skechers was weaker likely due to cannibalization from online sales.
15
Key Takeaways & Charts
Power of using multi datasets to generate 1 conviction signals
In the online channel, all brands accelerated y/y in April, with relatively high growth for Anta and Skechers. Adidas experienced the weakest growth in online sales.
Looking into Anta, there’s slight softness for Fila offline, offset by its strong sales online. Overall both Anta and Fila have outperformed the sector.
0%
20%
40%
60%
80%
100%
120%
140%
160%
Anta Fila Anta Fila
Offline (ex DepartmentStore)
Online
Anta and Fila growth in offline and online channel
4Q16 1Q17 Apr-May17
1Q17 y/y for online channel is estimated
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Adidas Anta Nike Skechers Overall
Sports Apparel Online Sales y/y
Mar-17 Apr-17
0%
10%
20%
30%
40%
50%
60%
70%
80%
Anta Nike Skechers Pou Sheng
Sports Apparel Offline (ex Department Store) Sales y/y
4Q16 1Q17 Apr-17
Sandalwood Advisors
Alpha from Data. Anta 72% Return
16
Sandalwood Advisors
How can you start using big data? What pitfalls to avoid?
1) Many funds start by throwing money at it, pitfall to avoid.
Many investors start by hiring a lot of data scientist – this is a big pitfall to avoid. While data scientists are useful, but many of them don’t have investment background, and can’t connect data with investment thesis.
2) Put someone senior with fundamental experience in charge of data. Don’t hire a junior, or market data role because they think data is too expensive, and very few people would listen to them even if they have great ideas.
3) Proof of concept. Look at your top 30 positions in your book, and ask which one can alternative data help you track better.
4) Interview as many big data firms as you can – ask them, how they made money in the past.
5) Try Structured data first.
6) Assign a time for analysts/pms to look at data. Most investors are stuck with traditional paradigms, and rarely assign time for data.
17
Sandalwood Advisors
Compliance - Must Ask before buying data
1) Are all your datasets anonymized, and aggregated?
2) Can you show me evidence you are authorized to sell the data?
3) Do you have a strong MNPI, Code of Ethics put in place?
4) Do you have exclusive clients? Are all funds allow to purchase your data?
5) Do you use expert network, or do they speak with company insiders to get info.
18
Sandalwood Advisors
Philosophy in using alternative data
• What is a hedge fund or active money management? When investors gives money, what are they really buying?
You are essentially buying a set of future decisions this investment managers make.A lot of things can influence the set of these decisions i.e. cognitive biases, divorce, their star analysts leaving.
So why not have a more systematic approach?
Investment is about seeing reality as clear as possible, then accurately projecting this ahead in the future.
Alternative data can help the investor see reality very clear.
19
Sandalwood Advisors
How will alternative data shape the financial markets?
1) Alternative data will make market more efficient.Efficient market hypothesis, more information = more accurate pricing. Now there is a more and more information available. Faster PC power, cheaper storage, and cheaper sensors.
2) Who would lose out? Tier 2-3 brokers as investor assign research budget away from them to big data firmsSmaller funds with no resources to consume or purchase data
3) Who are the winners? Funds who prepare now to understand data, especially the bigger funds who have resources to buy many datasetsBig data firmsThe economy, as capital gets allocated more efficiently.
20
Thank YouContact us: