Important investment information

Recommendation changes and distribution: Individual shares [PDF]

Recommendation overview: Individual shares [PDF]

Models

Quantitative recommendations for individual companies

Jyske Bank employs a quantitative model, Jyske Quant Global Large Cap Equities, to identify investment opportunities globally. Jyske Quant Global Large Cap Equities includes approx. 3,000 of the world's largest listed companies in terms of market value. Through a quantitative analysis the companies are scored on the basis of 23 key figures and subsequently ranked within three classic and well-documented factors:

• Value – the share is undervalued relative to the other companies.
• Quality – the share involves low risk and high quality relative to the other companies.
• Momentum – the share is attractive compared with other companies based on price development and analyst expectations.

The three factors are finally weighted equally, resulting in an overall Quant score, which forms the basis of the recommendation. The total Quant score ranks the companies within five quintiles. The Q1 quintile is the 20% best-scoring companies, i.e. the companies with the highest total score. On the other hand, the Q5 quintile is the 20% worst scoring companies etc. Jyske Quant Global Large Cap Equities is based on a thorough and well-documented method with strong historical results. The better the ranking, the higher the likelihood of obtaining a positive excess return.

The model excludes Financials and Utility since both sectors are considered being significantly different from other sectors. Among other things as a result of regulation and the companies’ typical capital structure. Also, there are large differences between the industries within these sectors. Jyske Bank will not make a fundamental assessment or a risk assessment of the company which is why we do not state a price target for the share.

Fundamental recommendations for individual companies

Jyske Quant Global Large Cap Equities (see description under the section ”Models”) is used to screen the market and identify future potential winners. We revise our recommendations on an ongoing basis. The analyst's assessment of the company is the determining factor behind a change of our recommendation from Buy or Strong Buy to Hold or Sell.

Jyske Bank employs one or more of the following models:

• Discounted cash flow model (DCF-model): In this model, the valuation of the company is based on the expected future free cash flows generated by the company's operations. The valuation is divided into a forecast period of ten years, an intermediate period, in which it is assumed that the company will be able to maintain its competitive advantage, and a terminal period. The value in the forecast period and in the intermediate period is calculated through a simple discounting of the free cash flows in the individual years with the relevant cost of capital, WACC. The value is calculated as:

  NPV of budget period (EV) = sum^T_(t=0) (FCF_t+1/(1+WACC)^(t+1))

  Where

  FCF_{t +1} = The free cash flow during period t+1 (both owners and lenders)
  WACC =The weighted average cost of capital (both owners and lenders)
  EV = Estimated value (both equity and interest-bearing debt)

• Dividend model: In individual cases, a dividend model is applied to determine the company's fundamental value. According to this model, the value of a company is the discounted value of all expected future dividend payments.
• Relative valuation: The fundamental value is compared with a relative valuation where key figures like P/E, P/B and EV/EBITDA are compared with those of the rivals. If the company is a consolidated company consisting of different business areas, a sum-of-the-parts valuation can be used where the valuation of the company is based on the value of the individual business areas determined on the background of key figures from comparable companies.
• Market sentiment: The recommendation and the price target are moreover adjusted for the expected news flow and the market sentiment based on knowledge of the industry and company-specific circumstances. This includes the momentum score in the Jyske Quant model. The momentum score comprises factors like earnings momentum, implicit volatility, analysts' estimate changes and recommendation changes.

Recommendations for individual companies: Trading Ideas

Jyske Quant Global Large Cap Equities (see description under the section “Models”) is used to screen the market and identify potential winners within the sector. In addition to Jyske Quant Global Large Cap Equities, the following methods are applied for screening, supplemented by the analysts’ in-depth industry knowledge. We continuously reassess our recommendations. The decisive factor for closing a trading idea will be the analyst’s assessment of whether the investment case remains intact, supported by input from the described models. When a trading idea is closed, it reflects the analyst’s view that the investment case has either been fulfilled or is no longer relevant, which is documented in a closure analysis. This means that Jyske Bank’s coverage of the stock ceases as of the trading idea’s closure analysis, and Jyske Bank will no longer provide any view on whether the stock should be sold, held, or bought going forward.

Furthermore, the trading idea is based on the following:

• Relative valuation: Key figures such as P/E, P/B, and EV/EBITDA are compared against the company’s own historical levels and those of the rivals. If the company is a consolidated company consisting of different business areas, a sum-of-the-parts valuation can be used where the valuation of the company is based on the value of the individual business areas determined on the background of key figures from comparable companies.
• Market sentiment: The trading idea is adjusted for the expected news flow and market sentiment based on knowledge of the industry and company-specific circumstances. This includes the momentum score in the Jyske Quant model. The momentum score comprises factors like earnings momentum, implicit volatility, analysts' estimate changes and recommendation changes.
• The following models may be applied if deemed value-adding:
  • Simplified DCF model: Based on consensus expectations for free cash flows over the next three years, discounted using the relevant cost of capital (WACC). A terminal value is then added, based on expected net income in year three multiplied by a P/E multiple derived from expected earnings per share, or alternatively an EV/EBITDA multiple.
  • Implicit growth requirement: This analysis estimates the annual revenue growth needed over the next 10 years to justify the current share price. The calculations are based on a Discounted Cash Flow (DCF) model, where free cash flows are discounted using the relevant cost of capital (WACC).

Strategy

Scenario model: The purpose of the model is to categorise expected return regimes in the equity market based on four simple and traditional market drivers: Valuation, interest-rate development, market sentiment and investor sentiment. The model has a clearly defined structure and is not based on distribution assumptions. Every week, the above-mentioned factors are evaluated, resulting in one of four expected return scenarios each with their own unique characteristics:

• Fundamentals: The most positive scenario with the highest expected return and normally distributed returns. Historically, approx. 62% of the observations have been positive in this scenario. This will typically occur around a market trough.
• Optimistic scenario: The second-best scenario is characterised by high expected returns, the lowest drawdown of all scenarios and a distribution of returns which is approximate normally distributed. Historically, approx. 56% of the observations have been positive in this scenario. This will typically occur between market trough and peak.
• Pessimistic scenario: The pessimistic scenario has a slightly negative expected return, the lowest upside of all scenarios and a distribution of returns that is leptokurtic and left skewed, i.e. more extreme and negative observations than you would have expected on the basis of a normal distribution. This will typically occur between market peak and trough.
• Speculative scenario: The scenario with definitely the lowest expected return and largest drawdown. The distribution of return is highly leptokurtic and left skewed, i.e. to a high degree more extreme and negative observations than you would have expected on the basis of a normal distribution. This will typically occur around market peaks.

For further details on our used models, click here [PDF].

Recommendation concepts

Quantitative recommendations for individual companies

The recommendation concepts Strong Buy, Buy, Hold and Sell are used. The recommendation alone depends on the quintile the company is placed in.

• Q1: STRONG BUY since there is a strong likelihood that the company will outperform the market.
• Q2: BUY since there is a moderate likelihood that the company will outperform the market.
• Q3: HOLD since it is likely that the company will perform in line with the market.
• Q4: SELL since there is a moderate likelihood that the company will underperform the market.
• Q5: SELL since there is a strong likelihood that the company will underperform the market.

Please note that the above descriptions of the recommendation concepts are average considerations. Consequently, it is still possible that companies in, for instance Q5, outperform the market whereas it is also possible that companies in Q1 underperform the market etc.

Fundamental recommendations for individual companies

Our recommendation structure consists of the four recommendations: Strong Buy, Buy, Hold and Sell. This results in the following recommendation structure:

Our recommendations are based on an evaluation of the forecast return over the next twelve months. The forecast return is the difference between the current price and our 12-month price target (our price target is inclusive of dividends over the period). The price target mirrors the potential we see in a share in the form of a price change and dividend within the coming 12 months.

It is the recommendation, not the price target, which is the anchor. A Buy recommendation will remain a Buy recommendation until changed, even if price increases have taken the price ‘too close’ to the price target. Hence, we will explicitly point out if we change our recommendation whereas a change of the price target will not necessarily trigger a new research report or comments.

Fundamental recommendations for individual companies: Trading Ideas

Our recommendation structure for trading ideas consists of the recommendation Buy while the trading idea remains active. When initiating a trading idea, our expectation is that the company’s price performance will be better than the average of its competitors (including dividends) over a horizon of 0–12 months

Strategy

Recommendations of risk level and asset allocation (overweight, underweight and neutral) are, among other things, assessed on the basis of Jyske Bank's expectation of whether an asset class will perform better, poorer or on level with the general market within the coming twelve months.

Recommendation changes

Models

Jyske Bank may apply one or more of the following models

Forecast model:

Jyske Bank uses a forecast model, which is adjusted to the individual issuer as needed so the model projects the factors relevant to the specific company in the specific sector. In the model, the items of the financial statements are projected to ensure a correlation between the historical financial statements and the projection of the model. Selected items of the income statement, the balance sheet and the cash flow statement are projected. The projected items form the basis of calculating credit-related financial ratios, which together with other relevant input is used to predict the future credit quality.

Expected credit spread:

Three models for expected credit spread are employed:

Expected changes in the credit quality – if the forecast model indicates a strengthening of the credit quality, it will point to a narrowing of the credit spread whereas expectations of a weakened credit quality will point to a widening of the credit spread.

Credit spread relative to comparable companies - identical issues should trade at identical credit spreads. Therefore the credit spread of an issue is compared with the credit spread of issues from comparable issuers. Same sector, rating, maturity, financial leverage, etc.

Credit spread relative to historical credit spread. The credit spread for an issue is compared with its history.

Recommendation concepts

In our analysis and screening process, individual companies are initially reviewed with a focus on their operational profile. If the overall operational profile is satisfactory, the company is referred to as a "Candidate," and its issuances can/should be included in a core portfolio (High Yield).

In evaluating the bond issuances of the companies (candidates), we utilize the recommendation concepts of Overweight (OW), Market Weight (MW), Underweight (UW), and Sell.

To receive an OW recommendation, a bond must, according to Jyske Bank's assessment, provide a return higher than comparable bonds within the same rating/sector over a 3–6-month period. The bond should be included in a core portfolio, and the timing is attractive considering either the expected credit development of the company and/or the current valuation of the bond.

To receive a MW recommendation, a bond must, according to Jyske Bank's assessment, provide a return at the same level as comparable bonds within the same rating/sector over a 3-6-month period. The bond should be included in a core portfolio, while the credit development of the company and/or the valuation of the bond are expected however to be neutral.

To receive an UW recommendation, a bond must, according to Jyske Bank's assessment, provide a return lower than comparable bonds within the same rating/sector over a 3-6-month period. The bond can be included in a core portfolio, but the timing is not optimal considering either the expected credit development of the company and/or the current valuation of the bond.

For companies where we determine that the operational profile no longer meets our criteria to be considered a Candidate, they will no longer be part of our analysis universe, and therefore, bonds from the company will receive a Sell recommendation. Coverage of these companies will continue for an additional 3 months.

Trading Size

Trading in corporate bonds can only take place in lots of minimum EUR 100,000. This means that corporate bonds can be traded in lots of, e.g., EUR 101,000 or EUR 102,000, but never in lots of, e.g., EUR 99,000. In the event of partial redemption or resale, there is a risk of ending up with a residual which cannot be traded in the market without a supplementary purchase.

Conflicts of interest

The analysts of Fixed Income Research may be physically placed next to the Sales and Trading departments. There may be consultations regarding publicly available information about the market, including prices, spread level and trading activity.

Research reports have been prepared by Research in Jyske Markets, which is a part of Jyske Bank A/S. Jyske Bank A/S is the owner of Jyske Realkredit. Jyske Bank A/S has considerable financial interests in relation to Jyske Realkredit.

Jyske Bank A/S is a joint owner of DLR Kredit and therefore has considerable financial interests in relation to DLR Kredit.

Jyske Bank participates in the market-maker scheme for Danish mortgage bonds and therefore in its capacity as this may have positions in these securities.

The following link shows the list of recommendations from Fixed Income Research [XLSX].

Models

Research reports

Danish mortgage bonds are modelled on the basis of an own-developed model based on RIO (mortgage model developed by Scanrate) consisting of a stochastic yield curve model to estimate future yield development and a statistic remortgaging model calibrated to borrowers' historical remortgaging behaviour.

In the preparation of research reports, the relative valuation, supply/demand, debtor structure and parameters from Jyske Bank's mortgage model, for instance, are given weight.

Recommendation

Jyske Bank applies one or more of the following models:

• Regression models with input from economic and financial data, including Jyske Bank's outlook on these (e.g. outlook on economic growth, inflation and monetary policy).
• Rate curve analysis, including slope of rate curve and curvature.
• Positioning, market psychological assessments, game theoretical scenario analysis and pattern recognition techniques are included in the assessment material.
• Danish mortgage bonds are modelled on the basis of an own-developed model based on RIO (mortgage model developed by Scanrate) consisting of a stochastic yield curve model to estimate future yield development and a statistic remortgaging model calibrated to borrowers' historical remortgaging behaviour. In the preparation of research reports, the relative valuation, supply/demand, debtor structure and parameters from Jyske Bank's mortgage model, for instance, are given weight.

For further details on our used models, click here [PDF].

Recommendation concepts

The research reports of Fixed Income Research focus mainly on two types of recommendations:

Relative recommendations focusing on relative value in the bond and derivatives markets. The interest and/or volatility risks are therefore typically hedged in other bonds or derivatives. This will often include both a BUY and a SELL recommendation.

Direct recommendations focusing on the absolute value in the bond market. The interest and/or volatility risks are therefore typically not hedged. This will often include either a BUY or a SELL recommendation.

Buy

The analyst estimates that the bond is relatively undervalued against comparable alternatives. It is the estimate of the analyst that the bond will generate a relatively better return than the alternatives within a horizon of typically three months.

Sell

The analyst estimates that the bond is relatively overvalued against comparable alternatives. It is the estimate of the analyst that the bond will generate a relatively poorer return than the alternatives within a horizon of typically three months.

Conflicts of interest

The analysts of Fixed Income Research may be physically placed next to the Sales and Trading departments. There may be consultations regarding publicly available information about the market, including prices, spread level and trading activity.

Research reports have been prepared by Research in Jyske Markets, which is a part of Jyske Bank A/S. Jyske Bank A/S is the owner of Jyske Realkredit. Jyske Bank A/S has considerable financial interests in relation to Jyske Realkredit.

Jyske Bank A/S is a joint owner of DLR Kredit and therefore has considerable financial interests in relation to DLR Kredit.

Jyske Bank participates in the market-maker scheme for Danish mortgage bonds and therefore in its capacity as this may have positions in these securities.

Models

Jyske Bank applies one or more of the following models:

• Regression models with input from economic and financial data, including Jyske Bank's outlook on these (e.g. outlook on economic growth, inflation and monetary policy).
• Rate curve analysis, including slope of rate curve and curvature.
• Danish mortgage bonds are modelled on the basis of an own-developed model based on RIO (mortgage model developed by Scanrate) consisting of a stochastic yield curve model to estimate future yield development and a statistic remortgaging model calibrated to borrowers' historical remortgaging behaviour. In the preparation of research reports, the relative valuation, supply/demand, debtor structure and parameters from Jyske Bank's mortgage model, for instance, are given weight.

For further details on our used models, click here [PDF].

Recommendation terms

The rates forecast reflects the expected direction of rates over a given horizon. The Rates forecast is an estimate of how in all probability we expect to see the rate at the horizon stated. The Rates forecast should therefore not be considered a precise minimum or maximum for the rate in the period given.

The interest rates and prices stated in the material are the latest observed in the market, cf. the time stamp. For information on mortgage bonds, please see jyskebank.dk/erhverv/ejendomsfinansiering/kurser. It is important to emphasise that the interest rates and prices stated are not tradeable for the client.

Models

Jyske Bank uses standard statistical tools and methods.

Financial exchange rate model for the US dollar

The model is a simple regression of financial time series to explain the development in the EUR/USD exchange rate. The purpose of the model is to look at the correlation between the development in the EUR/USD exchange rate and a number of explanatory financial variables that theoretically/intuitively make sense and which statistically have historically explained the development in the US dollar.

Fundamental exchange rate model for the US dollar

The model is a simple regression of fundamental time series to explain the development of the EUR/USD exchange rate. The purpose of the model is to look at the relationship between the development of the EUR/USD exchange rate and a number of explanatory macroeconomic variables that theoretically/intuitively make sense and which statistically have historically explained the development of the US dollar.

For further details on our used models, click here [PDF].

Recommendation concepts

Price targets reflect the direction in which the exchange rate is expected to move over a given horizon (direction and strength). The target is an estimate of how in all probability we expect to see the exchange rate at the horizon stated. The target should therefore not be considered a precise minimum or maximum for the exchange rate in the period given.

Models

Jyske Bank uses standard statistical tools and methods.

Fundamental Fair Value Brent oil price model

The model is a simple regression of fundamental time series to explain the price development of Brent crude oil. The purpose of the model is to look at the relationship between the development in the price of Brent crude oil and a number of explanatory fundamental variables that theoretically/intuitively make sense and that have statistically explained the development in the price of Brent crude oil.

Fundamental oil demand model

The model is a Vector Autoregression of fundamental cointegrated time series. The purpose of the model is to forecast global oil demand five years into the future based on forecasts of global economic growth and population growth.

For further details on our used models, click here [PDF].

Recommendation concepts

Price target

The price targets reflect the direction in which the  price is expected to move over a given horizon (direction and strength). The price target is an estimate of how, in all probability, we expect to see the price over the horizon stated. Therefore, the price target should not be regarded as an exact minimum or maximum for the price over the period given.

Recommendation

The individual client should together with his or her adviser always consider the recommendation and strategy selected relative to the client's risk profile and purpose.

Buy: For the period, believe in price increases that will offset costs (inventory, return on the investment, etc.); therefore exposure should be hedged over the period^*

Sell: May be caused by a belief that an actual price decline will take place, yet also that we do not see any possibilities of increases that can offset costs for inventory and return on the investment over the period^*.

Hold: We await more favourable opportunities for transactions at a later time, where we include reservations for inventory costs and liquidity.

^* for the 3 - 6-month term, our rule of thumb is that we recommend hedging of half of the full amount; for the 6 - 12-month term, we recommend a third of the full amount.

Models

The conclusions and results in this analysis are based on standard methods and standard models within macroeconomics and econometrics.

For further details on our used models, click here [PDF].