Observation:
Only two shares (Admiral and Burberry) have statistically significant difference with FTSE return. Their p-values are less 0.1 and t-Test values which exhibit distance from mean to 0 (in standard deviations) are more than 1.8.
Scatterplot mean returns vs. standard deviations
Observation
Investors seek high returns with lowest possible risk. Higher risk demands higher return. The most efficient shares generate high returns but with low risk. We have plotted the return of the selected shares and the market against their respective standard deviations (risk). Scatterplot shows two main groups of four shares each. GSK, HSBC, BP and MKS) have the lowest returns per unit of risk compared to the other shares (ADM, BTI, and BRBY) which generate significantly higher returns without proportionate increase in risk. It can be said that these last shares are more efficient.
Which one will we invest in?
With this in mind we recommend to invest in British American Tobacco for conservative investors and in Admiral and Burberry for risk-takers. As we have concluded before FTSE index has the lowest standard deviation (risk) and low return.
Stock and market returns: Single Regression
To analyse the connection between stock and market returns, as well as to build a model for predicting of future share return values and to understand how different stocks (industries) react on market fluctuations we have used regression analysis. The main results are provided in the following table.
Observation
First, there are no really strong linear regressions between shares and market. The highest adjusted R-square = 0.42 (BP). This means that maximum 42% of changes in returns can be explained by changes in market return. For this particular company such result could be obtained because of its substantial market capitalisation and huge impact on the index. The lowest value is 0.17 (Admiral) and it shows that changes in market have negligible influence on that share’s returns.
Second, we can see rather low p-values for beta in all equations. This implies that we have high confidence that the β <> 0. At the same time values of β are different: from 0.47 for GSK to 1.53 for Rio Tinto. Shares which have β less than 1 are more stable than the overall market and 1% of changes in market return leads to less than 1% in changes of particular shares. Four shares with β>1 (BTGroup, BP, Burberry and Rio Tinto) have higher fluctuations then the market.
Also, Beta concept in such regressions between shares and markets is very wide-spread in financial and investment analysis. It is used in CAPM-analysis for cost of own capital calculation, in Treynor-ratio for profit/risk analysis etc. We will use it for our decision as well.
Multiple Regressions,3
In order to explain (and be able to predict) the ‘behaviour’ of shares, we have tried to find some variables which influence our stocks. We have downloaded monthly prices of some commodities (oil, iron ore, cotton) as well as currency rates, and transformed them into monthly returns. We have added natural logarithm of market returns as well. We then started turning different combinations for multiple regressions and tried to find those with the highest adjusted R-square value and also all p-values are less than 0.05. Finally we checked it with “Stepwise” regression in StatTools. The results are:
Observation and Conclusions
- 5 shares have not shown any strong dependence on other variables besides the market. There could be two possible reasons: either market is the main force that influence them (but due to very low R-square this is unlikely) or we could not find an appropriate variable
- There are no shares which strongly interact with log (FTSE). This is also evident from the scatterplots
- For those shares which showed connection with other variables an overall model became more sustainable and its predictable force increased (R-square have increased on 0.05-0.16)
- 3 shares exhibited rather strong relationship with GBP/USD rate changes. Our reasoning is : 1) some companies operate mostly on USD market; 2) financial flows move between stock and FOREX markets depending on certain situations and significantly impact market prices
Portfolio analysis
To prepare better investment decisions we performed some portfolio analyses. We tried to find portfolios with better return/risk ratios then index and shares. For cumulative return and risk calculation we have used Marcowitz’s formulas for effective portfolios. We have done analysis of different pairs which have better returns and low internal correlation. As a result we have obtained four examples of such effective baskets.
Conclusions
Based on our analysis we have drawn the following conclusions:
- During the last 5 volatile years different shares and industries show distinguished results in terms of return and risk for their owners
- Investors should avoid GSK, HSBC, MKS and BP shares because of their negative returns.
- For conservative investors BTI (tobacco industry) is the best investment because it as moderate risk and higher than average market returns. At the same time we should keep in mind the GBP/USD rate as an additional influencing factor.
- For risk-takers it is better to invest in Admiral and Burberry keeping in mind high beta, and cotton and iron ore prices influence. The risk is higher but expected return is more than 2% per month and 15 times more than average market. Teynor ratio for these three shares is also the highest
- To decrease average risk our recommendation is to invest in portfolios which provide good returns and demonstrate low correlation with other shares, such as: ADM, BTI, BRBY and RIO.
Appendix I: Short summary of selected shares
BT Group: BT Group plc is a global telecommunications services company headquartered in London. It is one of the largest telecommunications services companies in the world and has operations in more than 170 countries. Through its BT Global Services division it is a major supplier of telecoms services to corporate and government customers worldwide. Its BT Retail division is a leading supplier of telephony, broadband and subscription television services in the UK, with over 18 million customers.
(Reference: http://en.wikipedia.org/wiki/BT_Group)
Admiral: The Admiral Group is a highly profitable and fast-growing financial services intermediary. Admiral was launched in 1993, with the aim of offering lower premiums to more people. It is the trading name of EUI Limited. In December 2009, the Group had thirteen brands, over 2.08 million customers insured and over 3500 members of staff worldwide. Group turnover, comprising total premiums written and other revenue rose 18% in 2009 to £1.08 billion (2008: £0.91 billion).
(Reference: http://www.admiralgroup.co.uk/)
BP: BP is one of the world's leading international oil and gas companies, providing its customers with fuel for transportation, energy for heat and light, retail services and petrochemicals products for everyday items. It has nearly 80,000 employees with Sales and other operating revenues of $297,107 million in 2010
(Reference: http://www.bp.com/)
British American Tobacco: BAT is the world’s second largest quoted tobacco group by global market share, with brands sold in more than 180 markets. BAT has more than 200 brands in their portfolio and holds robust position in each of their markets. In 2010, BAT reported revenue rise by 5 per cent to £14,883 million at current rates, while organic revenue rose by 3 percent at constant rates of exchange.
(Reference: http://www.bat.com/)
Vodafone: Vodafone Group plc is a global telecommunications company headquartered in London, United Kingdom. It is the world's largest mobile telecommunications company measured by revenues and the world's measured by subscribers (behind ), with around 341 million proportionate subscribers as of November 2010. It operates networks in over 30 countries and has partner networks in over 40 additional countries. It owns 45% of Verizon Wireless, the largest mobile telecommunications company in the United States measured by subscribers. It had a market capitalisation of approximately £93 billion as of 9 March 2011.
(Reference: http://en.wikipedia.org/wiki/Vodafone)
Mark and Spencer: Marks and Spencer plc (also known as M&S) is a British retailer headquartered in the City of Westminster, London, with over 700 stores in the United Kingdom and over 300 stores spread across more than 40 countries. It specialises in the selling of clothing and luxury food products.
(Reference: http://en.wikipedia.org/wiki/Marks_%26_Spencer)
Burberry: Burberry Group plc is a British luxury fashion house, manufacturing clothing, fragrance, and fashion accessories. According to Business Weekly, Burberry is the 98th most valuable brand in the world.
(Reference: http://en.wikipedia.org/wiki/Burberry)
Rio Tinto: The Rio Tinto Group is a diversified, British-Australian, multinational mining and resources group with headquarters in London and Melbourne. As of March 2009, Rio Tinto is the fourth-largest publicly listed mining company in the world with a market capitalization of approximately $34 billion.
(Reference: http://en.wikipedia.org/wiki/Rio_Tinto_Group)
HSBC: HSBC Holdings plc is a global banking and financial services company headquartered in Canary Wharf, London. As of 2011 it was the world's second-largest banking and financial services group and second-largest public company according to a composite measure by Forbes magazine. It has around 7,500 offices in 87 countries and territories across Africa, Asia, Europe, North America and South America and around 100 million customers. As of 30 June 2010 it had total assets of $2.418 trillion
(Reference: http://en.wikipedia.org/wiki/HSBC)
GlaxoSmithKline: GlaxoSmithKline plc (GSK) is a global pharmaceutical, biologics, vaccines and consumer healthcare company headquartered in London. It is the world's third largest pharmaceutical company measured by revenues (after Johnson & Johnson and Pfizer).
(Reference: http://en.wikipedia.org/wiki/GlaxoSmithKline)
Appendix II: Returns for the stocks and the market index
Table 1: Return Data for both market index and stocks
Appendix III: Scatterplot – stock vs. market
BT Group
Figure 1: Correlation - 0.503
Admiral
Figure 2: Correlation - 0.425
Vodafone
Figure 3: Correlation - 0.604
BTI
Figure 4: Correlation - 0.560
HSBC
Figure 5: Correlation - 0.557
M&S
Figure 6: Correlation - 0.465
Burberry
Figure 7: Correlation - 0.581
Rio Tinto
Figure 8: Correlation - 0.542
GlaxoSmithKline
Figure 9: Correlation - 0.445
BP
Figure 10: Correlation - 0.655
Appendix IV: Time series graph – stock vs. market
BT Group
Admiral
Vodafone
BTI
HSBC
M&S
Burberry
Rio Tinto
GlaxoSmithKline
B&P
Appendix V: Single Regression
BT Group
Admiral
Vodafone
BAT
HSBC
M&S
Burberry
Rio Tinto
GSK
BP
Appendix VI: Multiple Regression
BT Group
Admiral
Vodafone
BAT
HSBC
M&S
Burberry
Rio Tinto
GSK
BP
Appendix VII: Autocorrelation
Appendix VIII: ABBREVIATIONS
Single Regression Tables in Appendix V
Regression Tables in Appendix 2
3 We have not used lags because as per our analysis, provided in Appendix VII, there are no substantial autocorrelations
Return portfolio = Return share1 * Part share 1 (%) + Return share2 * Part share 2 (%)
St.Dev portfolio =